Eight women had painless thyroiditis, transient thyrotoxicosis, and low radioactive iodine uptakes but were without goiter; they constituted 15% of all thyrotoxicosis cases that we saw during the past year. Standard antithyroid antibody tests by tanned erythrocyte hemagglutination, complement fixation, and colloid and microsomal fluorescence, if present initally, were only weakly psotivie and became negative by 9 months. However, human antithyroglobulin antibody levels by a sensitive radioimmunoassay were elevated initially in all patients. In general, these radioimmunoassayable antibody levels fell, but they had not returned to normal by 9 months. Serial thyroid function tests and standard antithyroid antibody tests were most compatible with subactue thyroidtis. However, the persistence of radioimmunoassayable antithyroblobulin antibodies and recent reports of histologic evidence of lymphocyte thyroiditis in similar patients with goiter leaves open the possibility that this is a previously unrecognized presentation of chronic lymphocytic thyroiditis. Neither the cause nor the ultimate course of this syndrome is known.

Seven episodes of severe ketoacidosis in six nondiabetic patients were recognized at this hospital within an eighteen month period. All were women; one pregnant patient experienced two episodes at twenty-eight and thirty-two weeks' gestation. All patients admitted to heavy chronic alcohol intake and drinking binges. On admission, these patients were conscious and alert. Mean values were 143 mg./100 ml. for plasma glucose and 7.25 for arterial pH. Plasma bicarbonate was depressed with a mean anion gap of 18. Beta-hydroxybutyrate/acetoacetate ratio averaged 5.2. All patients had liver function abnormalities. Mean serum immunoreactive insulin was low, 5μU./ml. (n=2), while cortisol was markedly elevated at 76.5μg. per cent (n=3); mean growth hormone level was 14.1 ng./ml. (n=3). Free fatty acid concentration, measured on admission in one episode, was 1,945 mEq./L. Therapy with glucose, saline, and minimal amounts of alkali led to prompt recovery. Circulating levels of cortisol, insulin and growth hormone were measured serially in one patient during recovery; they quickly returned to normal. The dissociation of severe ketosis from glycosuria and hyperglycemia in these patients raises important questions concerning coupling of ketogenesis to gluconeogenesis. The striking preponderance of women, including one pregnant patient, reported with this condition also suggests a possible role for ovarian and placental hormones in its pathogenesis; fetal drain on carbohydrate reserves may further contribute to the tendency to ketosis. Alcoholic ketoacidosis may be relatively common, since we saw one case of this syndrome for about every four of diabetic ketoacidosis during this period.

The present study was designed to ascertain septohippocampal cholinergic alterations and their related behavioral deficits after early exposure to ethanol. Mouse pups were exposed to ethanol, 3 g/kg by daily subcutaneous injection on postnatal days 2-14. At age 50 days, the ethanol-exposed mice had significant reductions from control levels in eight-arm maze performance. For example, on the fourth testing day, the number of correct entries in the ethanol group was 21% below control levels (P < 0.05) and the number of trials needed to enter all arms was 48% above control (P < 0.001). It took the ethanol-exposed mice twice the time to reach criterion than it did control (P < 0.01). A 33% increase from control level in muscarinic receptor number (Bmax) was found in the treated mice of age 22 days and a 64% increase at age 50 days (P < 0.001). However, no differences between control and treated groups could be detected in the presynaptic component of the cholinergic innervation, choline acetyltransferase activity. The results suggest that early ethanol exposure acts on hippocampal function similarly to phenobarbital, probably via alterations in postsynaptic processes in the septohippocampal cholinergic pathways.

The characteristics of the glucose and insulin responses during the glucose tolerance test (GTT) in obese people as a group have not been established. We analyzed glucose and insulin levels during GTT in 160 healthy obese patients who averaged 42% over ideal body weight. Statistical upper limit of normal for 2-h glucose was 260 mg/dl in women and 206 mg/dl in men. Although there was a significant correlation between insulin and glucose levels in both sexes and between insulin and degree of obesity in women, r values were relatively low (r less than 0.4 for all). High insulin levels and delayed peak insulin were present in the majority of patients with normal GTT and absent in many of the most obese patients. Results indicate that upper limits of normal glucose for GTT in the obese are much higher than currently accepted criteria.