TWENTY years ago, O Keefe and Marks examined the

Transcription

1 X/98/$03.00/0 Vol. 83, No. 3 Journal of Clinical Endocrinology and Metabolism Printed in U.S.A. Copyright 1998 by The Endocrine Society Gin and Tonic and Reactive Hypoglycemia: What Is Important the Gin, the Tonic, or Both?* DANIEL FLANAGAN, PETER WOOD, ROBERT SHERWIN, KWASI DEBRAH, AND DAVID KERR Metabolism Unit, Royal Bournemouth Hospital, Bournemouth, England; Endocrinology Laboratory, Southampton General Hospital (P.W.), and the Division of Endocrinology, Yale University School of Medicine (R.S.), New Haven, Connecticut TWENTY years ago, O Keefe and Marks examined the metabolic consequences of drinking two or three large gin and tonics as a liquid lunch. The combination of gin and a mixer of regular tonic containing 60 g sucrose caused profound reactive hypoglycemia (mean blood glucose nadir, 2.7 mmol/l) up to 3 4 h afterward. Mixing alcohol with low calorie tonic or drinking regular tonic on its own had much less of an effect on blood glucose levels (1). The aim of this study was to test the hypothesis that hypoglycemia occurs not only as a result of hyperinsulinemia but also because of blunting of normal counterregulatory hormone release. Methods Eight healthy volunteers (five men and three women, aged yr), whose weekly alcohol consumption was less than 14 U, gave written consent for the study, which was approved by the local hospital ethics committee. After an overnight fast, 3 days of abstention from alcohol, and resting semirecumbent for 30 min, a retrograde, venous catheter was inserted into the dorsum of the nondominant hand with the hand placed in a hot box at 60 C to arterialize venous blood. The cannula was kept patent with a constant infusion of 0.9% saline. Using a randomized, double blind design, each subject participated in three studies separated by at least 2 weeks. They drank 1) 730 ml regular Indian tonic water ABSTRACT The objectives of this study were to test the hypothesis that alcohol can cause reactive hypoglycemia by attenuating the release of counterregulatory hormones. The subjects were eight healthy volunteers (five men and three women, aged yr). Each subject drank, using a randomized, double blind design 1) three large gin with regular tonics (0.5 g/kg alcohol and 60 g carbohydrate, mainly sucrose (G T); 2) the same amount of alcohol with Slim-line tonic (0.5 g carbohydrate; G alone); and 3) regular tonic without alcohol (T alone). Glucose, insulin, and counterregulatory hormone levels and middle cerebral artery velocity (MCAV), an index of cerebral blood flow, were measured. Alcohol levels averaged mg/dl. Peak insulin levels were similar in both studies in which regular tonic was consumed (95% confidence interval for difference, 6 to22 U/mL). After the ingestion of G T, the blood glucose nadir was lower compared to that with T alone (3.35 vs mmol/l; P 0.02) or G alone (3.35 vs mmol/l; P 0.01). After drinking gin, subjects reported typical hypoglycemic warning symptoms unrelated to the prevailing glucose level. In both alcohol studies, there was marked blunting of GH release (P 0.01). Despite a blood glucose nadir of 3.35 mmol/l, plasma epinephrine levels rose only slightly from 267 to 455 pmol/l (P NS) after G T. Ingestion of alcohol also caused a transient rise in right MCAV (P 0.05) followed by a late drop in velocity in both cerebral hemispheres in the G T study (P 0.05). In otherwise healthy individuals a combination of gin and regular tonic can induce reactive hypoglycemia. Acute ingestion of alcohol impairs the epinephrine response and markedly suppresses the release of GH in response to a fall in blood glucose levels. (J Clin Endocrinol Metab 83: , 1998) (Sainsbury, London, UK) containing 60 g carbohydrate mainly as sucrose plus 0.5 g/kg ethanol as Beefeater Gin equivalent to 97 ml for a 70-kg man (G T), 2) 730 ml Slim-line tonic (Sainsbury, London, UK) containing 0.5 g carbohydrate plus gin as described above (G alone), and 3) 730 ml regular Indian tonic water containing 60 g carbohydrate alone (T alone). Over the next 300 min, the following measurements were made at 30- to 60-min intervals: 1) blood alcohol by gas liquid chromatography; 2) right and left middle cerebral artery velocity, an index of cerebral blood flow, using a trans-cranial Doppler technique (SciMed, Bristol, UK) (2); 3) heart rate and blood pressure using an oscillometric sphygmomanometer; and 4) levels of plasma insulin and the counterregulatory hormones: catecholamines, GH, and cortisol. Plasma insulin [intraassay coefficient of variation (CV), 7.2% at 14.2 U/mL], cortisol (intraassay CV, 6.8 at 382 nmol/l), GH (intraassay CV, 6.2% at 7.8 U/L), and glucagon (intraassay CV, 10.9% at 40 ng/l) levels were measured by double antibody RIA. Catecholamines were measured by a radioenzymatic technique (Amersham, Arlington Heights, IL; epinephrine intraassay CV, 18% at 46 pg/ml; norepinephrine intraassay CV, 10% at 210 pg/ml). Symptoms of hypoglycemia were also determined. Subjects completed 100 mm visual analogue assessment of symptoms characteristically associated with hypoglycemia (facial flushing, palpitations, tingling, trembling, sweating, hunger, lightheadedness, weakness, anxiety, difficulty concentrating, and tiredness). Whole blood glucose was measured at 15-min intervals by a glucose oxidase method (Yellow Springs Instrument Co., Yellow Springs, OH; interassay CV, 4.0% at 4.2 mmol/l). Received September 4, Revision received November 13, Accepted November 18, Address all correspondence and requests for reprints to: Dr. David Kerr, Metabolism Unit, Royal Bournemouth Hospital, Castle Lane East, Bournemouth, Dorset, England BH7 7DW. diabet@rbch-tr. swest.nhs.uk. * This work was supported by grants from the Wessex Medical Trust and the National Institutes of Health (DK-20495). Statistical analyses Overall differences between serial measurements were examined by summary measures (3). Summary responses for each individual were calculated as the area under the curve. Contrasts in group means were compared by repeated measures ANOVA or, where appropriate, by paired Student s t tests. Where data were not normally distributed, comparisons were made after logarithmic transformation or Wilcoxon 796

2 ALCOHOL AND REACTIVE HYPOGLYCEMIA 797 signed rank tests. Results are expressed as differences between means and 95% confidence intervals (CI). Otherwise, data are shown as the mean se or median (interquartile range). Results Blood glucose, plasma insulin, and levels of counterregulatory hormone levels were similar at the start of each study. After ingestion of gin, blood alcohol levels rose rapidly to 60 5 and 66 6 mg/dl in the G T and G alone studies, respectively. As expected, blood glucose and insulin excursions were more substantial after drinking regular tonic (G T and T alone studies; Fig. 1). However, nadir blood glucose levels were lower after drinking G T compared to either G alone (3.35 vs mmol/l; mean difference, 0.60; 95% CI, 0.41 to 0.79 mmol/l; P 0.01) or T alone (3.35 vs mmol/l; mean difference, 0.52 mmol/l; 95% CI, ; P 0.02; Table 1). Insulin levels were markedly higher in both studies where regular tonic was consumed [peak level after G T, 51 vs. 17 U/ ml; after G alone, 34 U/mL (95% CI, 17 52; P 0.02); peak level after T alone, 43 vs. 17 U/mL; 26 U/mL (95% FIG. 1. Blood glucose (upper panel) and plasma insulin levels at baseline and over the subsequent 300 min. Results are shown as the mean SE. CI, 11 41; P 0.02)]. The rise in plasma insulin levels was similar in the G T and T alone studies (95% CI for difference, 6 to22 U/mL; P NS). Overall epinephrine, norepinephrine, cortisol, and glucagon responses were similar in each of the three studies, whereas plasma GH levels were markedly suppressed by alcohol on both occasions (Table 2 and Fig. 2). Despite a blood glucose nadir of 3.35 mmol/l, plasma epinephrine TABLE 1. Glucose nadirs for each individual for the three studies Subject no. G T (mmol/l) G alone (mmol/l) T alone (mmol/l) (225) 3.91 (240) 4.13 (180) (165) 3.94 (285) 3.28 (195) (210) 3.74 (195) 3.29 (195) (195) 3.83 (225) 3.91 (195) (210) 3.89 (210) 3.29 (195) (135) 3.9 (255) 4.17 (150) (90) 3.83 (210) 4.41 (195) (165) 4.5 (270) 4.8 (165) The time of glucose nadir in minutes from starting the drink is shown in parentheses.

3 798 FLANAGAN ET AL. JCE&M 1998 Vol 83 No 3 TABLE 2. Hormonal counterregulatory responses Area under curve Peak levels 95% CI for differences between studies Adrenaline (pmol/l) G T G T vs. G alone, 17 to 286 G alone G T vs. T alone, 85 to 279 T alone G alone vs. T alone, 37 to 112 Noradrenaline (pmol/l) G T G T vs. G alone, 49 to 177 G alone G T vs. T alone, 116 to 156 T alone G alone vs. T alone, 126 to 38 Glucagon (ng/l) G T G T vs. G alone, 10 to 15 G alone G T vs. T alone, 9 to17 T alone G alone vs. T alone, 9 to11 Cortisol (nmol/l) G T G T vs. G alone, 13 to 173 G alone G T vs. T alone, 18 to 168 T alone G alone vs. T alone, 66 to 76 Results expressed as the geometric means of calculated area under curves and peak hormone levels SE. Comparisons between studies are shown as 95% confidence intervals (CI) for differences between peak levels. GH values are shown in Fig. 3. FIG. 2. Plasma GH levels at baseline and over the subsequent 300 min. Results are shown as the mean SE. levels only increased from 267 to 455 pmol/l ( 178 pmol/l; CI, 20 to 395; P NS) and were unchanged from baseline in the other two studies. Levels of other counterregulatory hormones and norepinephrine were similar in all three studies. In both alcohol studies, there was a transient rise in right middle cerebral artery velocity followed by a late fall in velocity in both hemispheres in the G T study (Fig. 3). Drinking G T was also associated with a greater rise in heart rate [after G alone: 11.3 vs. 6.8; 4.5 beats/min (CI, ); after T alone: 11.3 vs. 5.1; 6.2 beats/min; (CI, ; both P 0.05)] and fall in mean arterial pressure [after G alone: 10 vs. 6; 4 mm Hg; (CI, 2 to 7); after T alone: 10 vs. 5; 5 mm Hg; (CI, 1 to 7; both P 0.05)]. After alcohol, the subjects reported an increase in dizziness, tingling, difficulty thinking, and sweating symptoms unrelated to the prevailing blood glucose level, with cumulative symptom scores greater in both alcohol studies (533 and 511 vs. 283 in the Talone study; P 0.01) Discussion Almost 20 yr ago, O Keefe and Marks reported that drinking a combination of gin and regular tonic (equivalent to three double measures) could cause reactive hypoglycemia many hours later, possibly as a result of direct priming of pancreatic -cells by alcohol (1). Here, we have confirmed their original observation that a combination of alcohol and a mixer containing a large amount of carbohydrate is more likely to cause reactive hypoglycemia than drinking either the gin or the mixer on its own. This did not appear to be due to greater hyperinsulinemia. In both alcohol studies, there was marked suppression of plasma GH levels when blood glucose declined below baseline. Drinking gin and regular tonic also caused an early transient rise in heart rate and fall in blood pressure and a later reduction in arterial velocity in both hemispheres. Previous studies have reported that acute and sustained alcohol use can suppress GH release in response to insulininduced hypoglycemia, arginine, and propranolol-glucagon (4 6). At night, acute and chronic alcohol adminis-

4 ALCOHOL AND REACTIVE HYPOGLYCEMIA 799 FIG. 3. Changes in right (upper panel) and left middle cerebral artery velocity. Results are expressed as the percent change SE from baseline for each study. *, P 0.05 for G T and G alone studies vs. T alone; #, P 0.05 for G T vs. G alone and T alone. tration is associated with a 75% reduction in the usual nighttime sleep-related release of GH (7). GH is an insulin antagonist; suppression may therefore lead to increased insulin sensitivity and a lower blood glucose level. Furthermore, after gin and regular tonic, plasma epinephrine levels did not rise significantly despite a blood glucose nadir below the blood glucose threshold for release of epinephrine (8). In healthy volunteers, it would be anticipated that there would be a 3- to 4-fold increase in epinephrine levels associated with the degree of hypoglycemia achieved here (9). However, we cannot entirely confirm the apparent lack of response, as comparable hypoglycemia was not achieved in each individual. After acute ingestion of moderate amounts of alcohol, most studies have shown an increase in heart rate, probably as a reflex response to arteriolar or venous dilatation (10). As noted by others (11, 12), acute ingestion of alcohol caused a transient rise in right middle cerebral artery velocity, an index of cerebral blood flow. The reduction in MCAV occurred around the same time as blood glucose levels fell below baseline and thus could be considered maladaptive, as it would be expected that cerebral blood flow would be maintained in the face of systemic hypoglycemia to maintain substrate supply to the brain. All three drinks contained quinine, which has been implicated in causing hypoglycemia (13), but the dose was small (59 mg/drink) and unlikely to be significant. In conclusion, drinking a combination of gin and regular tonic can cause a fall in blood glucose levels into the hypoglycemic range associated with suppression of GH release. It is possible that the involvement of alcohol in road traffic and other accidents may be a consequence of neuroglycopenia as well as a result of the well recognized direct effects of acute ingestion of alcohol on the brain. Acknowledgments The authors thank Aida Grossman for help with measurement of plasma catecholamine levels, and June Murphy for her assistance with the smooth running of the studies.

5 800 FLANAGAN ET AL. JCE&M 1998 Vol 83 No 3 References 1. O Keefe SJD, Marks V Lunchtime gin and tonic a cause of reactive hypoglycaemia. Lancet. 1: Aaslid R, Markwalder TM, Nornes H Noninvasive transcutaneous doppler ultrasound recording of flow velocity in basal cerebral arteries. J Neurosurg. 57: Matthews JNS, Altman DG, Campbell MJ, Royston P Analysis of serial measurements in medical research. Br Med J. 300: Priem HA, Shanley BC, Malan C Effect of alcohol administration on plasma growth hormone response to insulin-induced hypoglycaemia. Metabolism. 25: Kolaczynski JW, Ylikahri R, Harkonen M, Koivisto VA The acute effect of ethanol on counterrgulatory response and recovery from insulin-induced hypoglycaemia. J Clin Endocrinol Metab. 67: Redmond GP Effect of ethanol on endogenous rhythms of growth hormone secretion. Alcoholism Clin Exp Res. 4: Prinz PN, Roehrs TA, Vitaliano PP, Linnoila M, Weitzman Effect of alcohol on sleep and nighttime plasma growth hormone and cortisol concentrations. J Clin Endocrinol Metab. 51: Mitrakou A, Ryan C, Veneman T, et al Hierarchy of thresholds for activation of counterregulatory hormone secretion, symptoms and cerebral dysfunction. Am J Physiol. 260:E67 E DeFeo P, Gallai V, Mazzotta G, et al Modest decrements in plasma glucose concentration cause early impairment in cognitive function and later activation of glucose counterregulation in the absence of hypoglycaemic symptoms in normal man. J Clin Invest. 82: Stott DJ, Ball SG, Inglis GC, et al Effects of a single moderate dose of alcohol on blood pressure, heart rate and associated metabolic and endocrine changes. Clin Sci. 73: Wendt PE, Risberg J, Stenberg G, Rosen I, Ingvar DH Ethanol reduces asymmetry of visual rcbf responses. J Cerebral Blood Flow Metab. 14: Tiihonen J, Kuikka J, Hakola P, et al Acute ethanol-induced changes in cerebral blood flow. Am J Psychiatry. 151: White NJ, Warrell DA, Chanthavanich P, et al Severe hypoglycemia and hyperinsulinemia in falciparum malaria. N Engl J Med. 309:61 66.