sign in sign up
<<
Home , Reactive hypoglycemia

MEDICAL STAFF CONFERENCE

Reactive Mechanisms and Management

These discussions are selected from the weekly staff conferences in the Depart- ment of , University of California, San Francisco. Taken from tran- scriptions, they are prepared by Drs. Sydney E. Salmon-and Robert W. Schrier, Assistant Professors of Medicine, under the direction of Dr. Lloyd H. Smith, Jr., Professor of Medicine and Chairman of the Department of Medicine.

DR. SLEISENGER: * Grand Rounds today involves cose tolerance test. There was no known family an interesting disorder of homeo- history of diabetes. stasis-hypoglycemia. The patient with this syn- Results of physical examination, including neu- drome will be presented by Dr. Norman Coleman. rological testing, were entirely within normal lim- DR. COLEMAN:t A 26-year-old Caucasian woman its. The blood pressure was 120/70 mm of mer- was admitted to the hospital with a six-year his- cury and the pulse regul at 76 beats per minute. tory of which had increased in fre- Hemoglobin was 12.4 gra'ns per 100 ml and leu- quency over the nine months before admission. kocytes numbered 4,900 per cu mm with normal Although they lasted several days each time, the differential. Results of urinalysis, electrocardio- headaches were never' severe enough to cause her gram, x-ray films of chest and skull, and an elec- to stay home from work. They commenced in the troencephalogram were all within normal limits. neck, progressed to the side of the head and were A 24-hour urine measurement of 17-hydroxy- and not relieved by aspirin. Associated symptoms in- 17-ketosteroid excretion was normal. A morning cluded sleepiness, slurred speech, chills, tremu- plasma cortisol level was 44 micrograms per 100 lousness, occasional diaphoresis and , but ml, while an early afternoon level the same day no . The symptoms were unrelated to was 17 micrograms per 100 ml. Dexamethasone menses. There was no history of scotomata, loss (2 mg) was administered at 11 p.m. the same day, of consciousness, seizures or other neurological and the plasma cortisol level was 4.5 mg per 100 symptoms during these episodes. There was no ml on the following morning. The elevated plasma clear relationship to meals, although food did help cortisol was considered compatible with the his- the headaches. Past medical history included tory of oral contraceptive , since the pitui- some nervousness which had been treated with tary-adrenal axis showed normal suppression. The chlordiazepoxide hydrochloride (Librium®). The oral tolerance test showed the following patient had been taking norethindrane with me- results: stranol (Orthonovum®) birth control pills for the Time (hours) . 0 3 1 2 3 past year. A 32-year-old sister a 13 4 5 had history of Plasma glucose headaches and hypoglycemia documentgd by glu- (mg per 100 ml)...92 200 210 140 100 52 70 75 *Marvin H. Sleisenger, M.D., Professor of Medicine. tNorman C. Coleman, M.D., Assistant Resident in Medicine. At the third and fourth hour periods of the glu-

64 MAY 1971 * 114 * 5 cose tolerance test the patient complained of perinsulinism. While most readers in the scientific similar to that she had experienced in community would accept that Mr. Pezet's weak- the past. Reactive hypoglycemia was diagnosed ness, nervousness and could have been ex- and the patient was discharged with prescription plained by the reactive hypoglycemia, it was more of a high , low carbohydrate diet. On this difficult to accept the second half of the book dietary regime she has been asymptomatic for the wherein the cure of his asthma, hay fever and two months since discharge from the hospital. , as well as the cure of peptic ulcers, rheu- DR. SLEISENGER: This patient with hypoglycemia matic fever and arthritis in other patients, was presents a difficult problem in diagnosis and treat- attributed to stabilization of the blood level. ment. Dr. John Karam will discuss some of these Nevertheless, the book remains quite popular, de- aspects. spite its shortcomings, and a foundation located DR. KARAM:* Patients who have rather nonspe- in New York distributes it nationally to its mem- cific complaints and no abnormalities on physical bers who suffer in common from hypoglycemia. examination are always a difflcult management In the patient presented today the obvious ques- problem. A succession of normal laboratory tests tion is whether the headaches, nervousness and usually confirms the initial impression that or- other nonspecific symptoms had any correlation ganic disease is not present, and the symptoms are with the documented measurement of a plasma then attributed to the tensions of modem living. sugar level of 52 mg per 100 ml recorded three Usually the only form of therapy given is reassur- hours after 100 grams of glucose was administered ance, and since it does not often succeed the pa- orally. Many do not consider a blood tient often consults a number of physicians and glucose level above 40 mg per 100 ml or a plasma tries a succession of tranquilizers without benefit. glucose above 46 mg per 100 ml to be low enough Inevitably the question of hypoglycemia is raised, to cause hypoglycemic symptoms. In one study, frequently by the patients themselves owing to continuous automatic monitoring of blood glucose wide publicity in the lay press. "Health food" demonstrated that nine out of twelve asympto- store shelves are lined with current books para- matic subjects had a blood glucose between 44 phrasing the original classic, Body, Mind, and and 58 mg per 100 ml at some time during an oral Sugar,' which emphasizes an almost mystical role .2 However, the glucose re- of low blood sugar in many of the physical ills of sponse of the present patient must be considered our times. This book, written in 1951 and unre- in context with the clinical problem. The general vised since the original publication, has been re- pattern of the oral glucose tolerance test was ab- printed almost every year and remains a best normal, with initial hyperglycemia (200 mg per seller. The authors of this book are a , 100 ml in the first half hour) and a relatively rapid E. M. Abrahamson, and one of his patients, A. W. fall in plasma glucose to 52 mg per 100 ml by three Pezet, who had multiple complaints including hours. As continuous monitoring was unavailable, nervousness, weakness, and fatigue which per- we cannot ascertain the exact minimal level to sisted despite attempts at management by numer- which plasma glucose may have fallen in this pa- ous physicians. It was only after Dr. Abrahamson tient. The development of her chief complaint, performed a glucose tolerance test and demon- headache, at the time of the lowest measured strated a reduced blood sugar at three to four plasma sugar level and her own observation that hours that sucoessful therapy was begun. This food often relieved her headaches were additional therapy consisted of frequent, small feedings of a points suggesting an etiological relationship be- low carbohydrate, high protein diet. With the tween symptoms and blood glucose levels. help of his patient, who was a professional jour- The pattern of response to the oral glucose tol- nalist, Dr. Abrahamson wrote a very entertaining erance test is a great help in classifying patients and expressive story of the dramatic results ob- with hypoglycemia.3 The presence of a "reactive" tained in this and other cases. rather than a "fasting" hypoglycemia in this pa- The first half of the book gave an excellent and tient makes it unlikely that the diagnosis is the remarkably accurate scientific description of car- result either of autonomous production of bohydrate with emphasis on the role by an or of heterogenous factors in- of insulin in its regulation and the hazards of hy- volved in the hypoglycemia related to nonpancre- *John H. Karam, M.D., Assistant Professor of Medicine. atic tumors.

CALIFORNIA MEDICINE 65 The Western Journal of Medicine REACTIVE HYPERINSULINISM

250 - ACCELERATED (alimentary) -DELAYED (diabetic) ,/"E ~~~~(a) 150

E 50

E 250 E SUSTAINED (obese nondiabetic) SUSTAINED (pseiudo dia'be-tic) E 150 a _ \ z ~~~~(c) .fftf-\ftf.. ui -Ij

LI) SO z -J WITHOUT HYPERINSULINISM 0V . ,XCOUNTER REGULATORY DEFICIENCY IDIOPATHIC I,v LJ ,___ (f) LI) 0 50 4 2%3%4 - co eI I 1 2 3 4 S 1 2 3 4 5 TIME IN HOURS Chart 1.-Patterns of Reactive Hypoglycemia: Reactive hypoglycemia with reactive hyperinsulinism (a,b,c,d). Reactive hypoglycemia with normal insulin response (e,f). The abscissa records time after ingestion of 100 gm of glucose.

Classification of Reactive Hypoglycemia TABLE 1.-Classification of Reactive Hypoglycemia With the establishment of a "reactive" type of "Early" Hypoglycemia hypoglycemia by clinical history, as well as by Postgastrectomy plasma glucose measurements, it is then useful to Increased vagal tone determine whether the pattern of response to the "Late" Hypoglycemia oral glucose load is one of "early" or "late" hypo- Diabetic reactive hypoglycemia glycemia. A typical "early" hypoglycemia occurs Nondiabetic 'late" hypoglycemia Obesity within three hours of the ingestion of glucose, Pseudodiabetes whereas 'late" hypoglycemia occurs three to five Counter regulatory deficiency hours after receiving the glucose. The patho- Idiopathic reactive hypoglycemia genetic mechanisms and management of these re- sponses are quite different. "Early" Hypoglycemia. The present patient though serum insulin levels were not measured demonstrated a pattern of glucose response in the in this patient, they almost invariably reach very first hour that is characteristic of the alimentary high levels in the first hour when associated with type (Chart 1,a). There was an exaggerated rise this particular type of plasma glucose response. of the ascending limb of the curve to reach supra- "Early" hypoglycemia is quite common in post- normal values within 30 minutes, followed by a gastrectomy cases4 but may also occur in other rapid descent associated with symptoms. Al- patients, especially in thin, nervous, compulsive,

66 MAY 1971 * 114 * 5 hyperactive persons who often complain of gas- TABLE 2.-Historical Sequence of Development5 of Concept that Betacytotropic Factors Are Present in trointestinal difficulties compatible with increased Normal Persons vagal tone. 1904 Secretin Induced B-Cell Hyperplasia The pathogenesis of this pattern of response 1923 Secretin Reduced Blood Sugar in Dogs consists of at least two components: accelerated 1940 Hypoglycemic Effect of Secretin Not Confirmed 1964 Oral Glucose Increases Insulin Levels More Than rate of glucose absoirption and intestinal betacy- IV Glucose totropic factors. 1966 Extracts of Duodenum Increased Blood Insulin Level, as did the following: Accelerated Glucose Absorption. The most ob- Secretin vious cause of accelerated glucose absorption is Pancreozymin an abnormally rapid absorption of the alimentary Castrin glucose in patients who have had gastrectomy, as -Like Substance well as in those who simply have accelerated gas- tric emptying, possibly owing to increased vagal factors are present in normal subjects. In 1904 it tone. Hyperinsulinism, resulting from the initial was reported that secretin administration pro- hyperglycemia, as well as from the increased duced hyperplasia. This finding was vagal effect on pancreatic beta cell secretion, pro- overlooked until after the discovery of insulin, at duces hypoglycemia. which time it was reported that secretin also re- Intestinal Betacytotropic Factors. It has been duced the blood glucose of dogs. However, when clearly demonstrated that the hyperinsulinism improved micromethods were developed for meas- which is observed in patients who have had gas- uring blood glucose, it was observed and subse- trectomy cannot be the result of only the accel- quently confirmed that there was no real hypo- erated rate of glucose absorption, since similar glycemic effect of secretin in dogs.5 Thus, for a hyperinsulinism does not develop when glucose time the concept of a role of secretin as a beta- is infused rapidly to produce comparable levels cytotropic substance was discarded. Only after of hyperglycemia.4 This finding, that the insulin insulin measurements demonstrated the superior response depends on the route of glucose admin- insulinotropic effect of oral as compared with in- istration rather than on the circulating glucose travenous glucose was the problem reinvestigated. level alone, was a very exciting development in When extracted duodenal mucosa was shown the understanding of insulin secretion and its con- to stimulate insulin release6 the term "gut factor" trol. It had long been known that a given amount became widely used, and its chemical isolation, of glucose produced higher blood glucose levels purification, and characterization were expected when administered intravenously than when in- momentarily. The early reports of secretin-in- gested. The assumption was that when glucose duced beta cell hyperplasia seemed to be vindi- was administered orally the time necessary for cated when secretin was shown to stimulate in- gastrointestinal absorption and removal by the sulin secretion both in vivo and in vitro, despite liver before reaching the systemic circulation ac- its verified lack of hypoglycemic effect in animals. counted for the lower levels. If this were the case, However, investigators were dismayed when in- then the higher glucose concentration in the pan- stead of finding a single betacytotropic factor, creatic artery after intravenous administration they found that just about every gastrointestinal would have been expected to produce higher hormone tested had insulinotropic properties.7 levels of circulating insulin than did glucose by Unger8 reported finding a variety of different pep- mouth. However, when techniques were devel- tides with glucagon-like immunoreactivity in gas- oped to measure insulin, it was surprisingly re- trointestinal mucosa; these peptides can enhance ported that oral glucose produced much higher insulin release. Other known insulin secreto- insulin levels than did comparable intravenous gogues found in gostrointestinal mucosa include doses.5 This finding suggested that insulin release gastrin and cholecystokinin-pancreozymin.7 The was not dependent on blood glucose levels alone question of the nature of this gut factor is still and that something produced during alimentation unresolved. It is not clear whether all of these was enhancing the stimulatory effect of glucose enteric hormones participate in the enhancement on insulin release. of insulin release by some common mechanism Table 2 relates the historical sequence5 in the such as cyclic-AMP (Adenosine 3' 5' monophos- development of the concept that betacytotropic phate) generation, or whether, as Creutzfeld7

CALIFORNIA MEDICINE 67 The Western Journal of Medicine speculates, the available data represents pharma- cytropic factors, contribute to the well-docu- cological observations; and the actual physiologi- mented islet cell hyperplasia and sustained hy- cal betacytotropic hormone remains as yet un- perinsulinism found in most obese persons.12 detected. When islet cell hyperplasia progresses so that the Regardless of its nature, an enteric insulino- postprandial hyperinsulinism overcompensates tropic factor certainly seems to exist and may well for the peripheral antagonism, "late" hypogly- be produced excessively in postgastrectomy pa- cemia occurs. tients and other persons in whom gastric contents Nondiabetic 'late" hypoglycemia has also been empty at an accelerated rate. An increase in glu- described in patients who have been referred to cagon-like immunoreactivity after oral glucose as having pseudodiabetes (Chart 1,d). These has been reported recently in two patients with patients have an early and sustained hypersecre- reactive hypoglycemia and has been suggested as tion of insulin in association with both an ab- a pathogenetic factor in their hypoglycemia.9 normal elevation and a delayed clearance of glu- "Late" Hypoglycemia. Diabetic reactive hypo- cose. In these patients mild carbohydrate intol- glycemia (Chart 1,b) was first described in 1956 erance occurs during the initial two or three by Seltzer et al'0 in a classic study of 110 patients hours of a glucose tolerance test, and then a with normal fasting blood sugar but impaired car- "late" hypoglycemia occurs at four or five hours. bohydrate tolerance. Hyperglycemia persisted The presence of a normal or, more often, a for as long as two to three hours after ingestion supranormal early phase of insulin release differ- of glucose, but then was followed by a transient entiates these patients from those with "true" dia- fall to hypoglycemic levels between the third and betes mellitus. In pseudodiabetes an early re- fifth hour. lease of insulin occurs in response to glucose in- In contrast to hyperactive patients with "early" gestion. An end organ resistance to the action of hypoglycemia, patients with "late" hypoglycemia these high levels of circulating insulin, however, tend to be more phlegmatic, with a high incidence provides the best explanation for the abnormally of obesity, and frequently have a positive family elevated blood glucose levels during the first history of diabetes mellitus. Serum insulin re- hour of the glucose tolerance test. sponse to glucose is delayed initially in patients In a series of 238 obese subjects reported upon with early adult-onset diabetes, accounting for by Faludi et al,'3 101 had "late" hypoglycemia their carbohydrate intolerance, but gradually rises four hours after a glucose tolerance test. Of so that by the second and third hour of the glu- these, 27 had normal carbohydrate tolerance in cose tolerance test hyperinsulinism occurs." Thus, the first three hours, while 74 showed minimal the pathogenesis of this type of "late" hypogly- hyperglycemia during this time. Although meas- cemia is believed to be within the beta cell itself, urement of early insulin release was not reported and it is thought to represent one of the earlier in this study, it is our experience that the major- manifestations of diabetes mellitus. ity of obese subjects with normal fasting blood Nondiabetic "late" hypoglycemia has been de- glucose and only minimally impaired carbohy- scribed in obese patients who have an early and drate intolerance would have a supranormal ini- sustained release of insulin in association with tial release of insulin after glucose and could be peripheral insulin antagonism (Chart 1,c). This termed as having "pseudodiabetes."l2 peripheral antagonism has been variously ascribed Certain regulatory mechanisms are normally to abrnormally enlarged adipose cells, intramus- available to counteract any excessive hypogly- cular inhibition of glucose metabolism by raised cemic effect of insulin during the latter part of a fatty acids, or a combination of these and other, glucose tolerance test (Chart 1,e). These include as yet unknown, antagonists. Whatever the cause, pancreatic glucagon, catecholamineps, growth this peripheral antagonism to insulin's action re- hormone, sympathetic nervous stimulation, and sults in a delayed postprandial clearance of circu- cortisol. Deficiency in release of these substances lating glucose, although an abnormal elevation of in response to hypoglycemia may thus lead to blood glucose does not occur. The prolongation more pronounced and sustained hypoglycemia. of glucose clearance may stimulate the beta cells In the presence of normal glucose tolerance and, in addition to the excessive ingestion of and normal insulin release over the first two calories with their consequent release of beta- hours, patients with reactive hypoglycemia should

68 MAY 1971 * 14 * 5 first of all be carefully evaluated for deficiencies, Delayed hyperinsulinism with "late" hypogly- of the counter regulators. In the absence of cemia (Chart 1,b) is the pattern of reactive hy- these deficiencies, such cases have been classi- poglycemia found in patients with diabetes mel- fied as idiopathic reactive hypoglycemia (Figure litus. Elimination of sugar-containing foods and l,f). frequent feedings with increased proteins re- duces the initial hyperglycemia that precipitates of the delayed hyperinsulinism. Also, in this situa- Diagnosis and Treatment tion restoring the early phase of insulin release Reactive Hypoglycemia with tolbutamidel- administration has been re- In patients without fasting hypoglycemia who ported to be clinically useful.'8 have vague autonomic symptoms temporally re- In sustained hyperinsulinism associated with lated to meals, diagnosis is best established by a obesity and pseudodiabetes (Chart 1,c and d) five-hour oral glucose tolerance test. This test caloric intake is lowered to reduce the obesity is preferably supervised by a physician so that, and interrupt the cycle producing islet cell hy- in addition to regular blood sampling every 30 perplasia and consequent hyperinsulinism. SmalL minutes, he can obtain a blood sugar determina- spaced feedings low in carbohydrate, free of tion at the first sign of symptoms, lest a transient sugar and increased in protein are also beneficial. period of hypoglycemia be missed. In addition, phenformin hydrochloride (DBI®) Treatment is based on the pattern of reactive has proved to be effective in a large group of hypoglycemia obtained. The measurement of obese patients by abolishing late hypoglycemia serum insulin and blood sugar better delineates following a glucose load.'3 The mechanism of the physiological disturbance as shown in Chart action involved with the administration of phen- 1. In our laboratory, fasting levels of insulin sel- formin-may be a reduction in the rate of absorp- dom are above 30 ,lu per ml in normal persons tion of ingested glucose by the intestinal mucosa, but may be as high as 50 ,uu per ml in obese per- which results in a smaller elevation of blood glu- sons. After a glucose load, the normal person cose and consequently decreased insulin produc- usually does not have an increase in serum insulin tion by the hyperplastic pancreatic beta cells. level above 100 juu per ml during his peak insulin Specific replacement therapy is difficult in most response between 30 and 60 minutes. counter regulatory deficiencies, except in the case In an accelerated hyperinsulinism with early of cortisol insufficiency. Thus, reliance on multi- hypoglycemia (Chart 1,a) treatment is aimed at ple, small, spaced feedings to minimize increases reducing the large quantity of food suddenly be- in blood sugar and insulin secretion would be rec- ing presented to the small intestine after meals ommended for these patients as well as for those in patients who have had gastrectomy or those with idiopathic reactive hypoglycemia. with increased vagal tone, in order to reduce It is of interest that most patients in all the cate- rapid excursions of blood glucose. The patient gories of reactive hypoglycemia seem to respond is offered small, frequent, high protein feedings to a substitution of protein for carbohydrate in which contain no processed sugar and only lim- their diet even though protein ingestion results in ited carbohydrate. Anti-cholinergic drugs have insulin release. It has been suggested that the also been used to inhibit vagal tone and delay concomitant production of glucagon offsets the gastric emptying. In a well studied case, such action of this insulin.'9 cholinergic blockade reduced the rate of gastric In summary, the patient discussed today, who emptying, normalized the exaggerated hyperin- is not obese, has a glucose pattern suggestive of sulinism and corrected the "early" hypoglyce- alimentary hypoglycemia with an exaggerated mia.'4 Certain pharmacological agents, including early rise and then a return to normal and finally diazoxide,14 mannoheptulose,'5 and diphenylhy- hypoglycemic levels. The Orthonovum® therapy dantoin sodium (Dilantin® ),16 may induce hy- may have produced just enough insulin antagon- perglycemia in man with a reduction in serum ism to slightly delay the hypoglycemia in this pa- insulin levels. While these agents offer possible tient from the 120 to 150 minutes normally seen therapeutic approaches in this and other types of in the more classical cases of alimentary hypogly- reactive hypoglycemia, they as yet have not re- cemia. Unfortunately, insulin measurements were ceived adequate clinical trials. not undertaken to confirm the impression of accel-

CALIFORNIA MEDICINE 69 The Western Journal of Medicine erated hyperinsulinism. Her clinical picture is, atypical symptoms which occur during docu- however, typical enough to suggest a trial of low mented reactive hypoglycemia and are relieved carbohydrate, high protein diet of small, spaced by eating. feedings. Since the hypoglycemia and symptoms are rather mild, this therapy should be adequate, TRADE AND GENERIC NANIES OF DRUGS with anticholinergic agents kept in reserve in case Libriums'...... chlordiazepoxide hydrochloride symptoms become more severe. Phenformin might Orthonovum ..norethindrane with mestranol also be of use in alimentary hypoglycemia, since Dilantin® diphenylhydantoin sodium it reduces the rate of glucose absorption. DBI.. phenformin hydrochloride DR. HAVEL:* How frequently do atypical symp- toms, such as headache, occur? Since the glucose REFERENCES 1. Abrahamson EM, Pezet AW: Body, Mind, and Sugar. New York tolerance test alone is not very reliable as a diag- City, Holt, 1951 2. Burns TW, Bregant R, Van Peenan HD, et al: Observations on nostic aid, and not everybody can measure insulin blood glucose concentration of human subjects during continuous sam- levels, how much do you rely on the history in pling. Diabetes 14:186-193, Apr 1965 3. Anderson JW, Herman RH- The classification of reactive hypo- evaluating the nature of the disorder? glycemia. Amer J Clin Nutr 22:646-649, May 1969 4. Holdsworth CD, Turner DS, McIntyre N: Pathophysiology of DR. KARAM: This question is difficult to answer. post-gastrectomy hypoglycemia. Brit Med J 4:257-259, Nov 1, 1969 5. McIntyre N, Holdsworth CD, Turner DS: Intestinal factors in As physicians we were taught that if a person's control of insulin secretion. J Clin Endocr 25:1317-1324, Oct 1965 blood sugar drops from a high level to a low level, 6. Dupr6 J, Beck JC: Stimulation of release of insulin by an extract of intestinal mucosa. Diabetes 15:555-559, Aug 1966 we could expect certain autonomic symptoms at- 7. Creutzfeldt W, Feurle G, Ketterer H: Effect of gastrointestinal hormones on insulin and glucagon secretion. New Eng J Med 282:1139- tributed to epinephrine. In Body, Mind, and 1141, May 14, 1970 8. Unger RH: Enteric and other humoral agents affecting insulin Sugar, the authors describe the usual responses, secretion. 9th Research Symposium of American Diabetic Association, but claim that many atypical symptoms can also Colorado Springs, 1970 9. Rehfeld JF, Heding LG: Incieased release of gut glucagon in re- occur. They describe, for instance, the pitcher in active hypoglycemia. Brit Med J 2:706-708, June 20, 1970 10, Seltzer HS, Fajans SS, Conn JW: Spontaneous hypoglycemia as the baseball game who had perfect control until an early manifestation of diabetes mellitus. Diabetes 5:437-442, Nov/Dec 1956 the eighth inning, when he suddenly blew up and 11. Yalow RS, Berson SA: Plasma insulin concentrations in non- diabetic and early diabetic subjects. Diabetes 9:254-260, July/Aug walked everybody. They also imply that most 1960 husbands who argue with their wives before sup- 12. Karam JH, Grodsky GM, Forsham PH: Insulin secretion in obesity: Pseudodiabetes. Amer J Clin Nutr 21:1445-1454, Dec 1968 pertime are "hypoglycemic," and if only their eat- 13. Faludi G, Bendersky G, Gerber P: Functional hypoglycemia in early latent diabetes. Ann NY Acad Sci 148:868-874, Mar 26, 1968 ing habits were altered their marriage would be 14. Veverbrants E, Olsen W, Arky RA: Role of gastrointestinal fac- blissful. Without documentation of blood glucose tors in reactive hypoglycemia. Metabolisni 18:6-12, Jan 1969 15. Lev-Ran A, Laor J, Vins M, et al: Effect of I.V. D-Mannohep- levels, it is hard to prove (or disprove) whether tulose on blood glucose and insulin levels in man. J Endocr 47:137- 140, May 1970 these are symptoms of hypoglycemia. However, 16. Hofeldt F, Levin S: Effectiveness of dilantin in reactive hypo- glycemia. Unpublished data I would say that I believe atypical symptoms, such 17. Varsano-Aharon N, Echemendia E, Yalow RS, et al: Early in- as headache, can occur; but I do not know how sulin responses to glucose and to tolbutamide in maturity-onset dia- betes. Metabolism 19:409-417, June 1970 I could be certain. I suppose we should give the 18. Economou PG, Ryan WG, Schwartz TB: Mild diabetes and prediabetes: Response of the hypoglycemia to tolbutamide. Presby- patient the benefit of the doubt if we observe terian-St. Lukes Hospital Medical Bulletin 4:46-52, Apr 1965 19. Freinkel N, Metzger BE: Oral glucose tolerance curve and hypo- *Richard J. Havel, M.D., Professor of Medicine. glycemias in the fed state. New Eng J Med 280:820-828, Apr 1969

70 MAY 1971 * 114 * 5