We analyzed the causes of death in 74 non-insulin-dependent diabetes mellitus (NIDDM) patients who died in Koseiren Takaoka hospital between 1982 and 1988 and compared with 334 non-diabetic dead patients.
The first cause of death in NIDDM was malignant neoplsma (35.1% of NIDDM). The second was ischemic heart disease (17.6%). The third was infection (12.2%). The ratios of cerebral infarction and diabetic nephropathy were 9.5% each. The ratios of ischemic heart disease and cerebral infarction in diabetics were twice those in non-diabetics. The tratio of uremia in diabetics wassix-fold. In patients over 61 years old, the ratios of ischemic heart disease and diabetic nephropathy were large. Nevertheless, in patients under 60 years old, the proportion of cerebral infarction was higher. The proportions of cerebral infarction and ischemic heart disease were higher in the group of long diabetic duration than in group of short duration.
Compared with past Japanese reports, the proportions of malignant neoplasma and ischemic heart disease in diabetics increased and the proportions of diabetic nephropathy and coma decreased.
This study concluded that not only the control of diabetes mellitus but also the examination of malignant neoplasma was important in management of diabetes mellitus. The proportion of the causes of death in diabetics will change with changes of the circumstances and the progress of medical treatment.

The patient was a 56-year-old male. He had a history of alcoholism since 1975 and diabetes mellitus since 1978. He was treated with insulin therapy. But the control of diabetes mellitus was very poor and he had six hypoglycemic comas attacks. He had diabetic triopathy but no symptoms of gait disturbance, dementia and incontinence. Brain computerized tomography and magneting resonance imaging revealed severe communicating hydrocephalus with mild brain atrophy.¹¹¹In-DPTA cisternography revealed retension of isotope in the ventricle after 48 hours. The pressure of cerebrospinal fluid was normal.
This case report is interesting as it suggests the relationship between normal pressure hydrocephalus and hypoglycemia.

To determine risk factors for progression of diabetic retinopathy, a study was conducted in 92 patients with diabetes whose HbA1c levels were more than 9.5% when they were admitted to our hospital.
The progression of retinopathy was observed in 50% of the patients during 12 months after admission. Their cases were diagnosed as preproliferative retinopathy. The duration of no-treatment, serum cholesterol level, blood pressure and proteinuria were significantly greater in the deterioration group of retinopathy than in the no-deterioration group. Nerve conduction velocity was significantly lower in the deterioration group of retinopathy than in the no-deterioration group. The reduction of HbA1c level during the period of 3 months after admission was significantly higher in the deterioration group than in the improved group. In the patients whose initial levels of HbA1c were less than 10%, the deterioration of retinopathy was not observed. In conclusion, these findings suggest that the duration of no-treatment, serum cholesterol level, blood pressure, proteinuria, nerve conduction velocity, retinopathy severity, the HbA1c level at admission and the reduction of HbA1c level during the 3-month period are risk factors for the progression of retinopathy.

A 62-year-old woman was adimitted to our hospital because of thirst and a body weight loss of 9kg. She had a history of vasospastic breast pang 5months before. The level of fasting plasma glucose was 320 mg/dl and the level of HbA_1c was 13.0%. The autoantibody to glutamic acid decarboxylase was positive. The level of urine C-peptide was 28.6μg/day. She rejected insulin injections and was treated with glibenclamide. The level of urine C-peptide increased to 70.0μg/day. The disease was controlled with the levels of 6.0% of HbA_1c. But her condition became uncontrollable gradually after 6 months and she was treated by insulin therapy.
Insulin secretion in type 1 diabetes mellitus was transiently improved by sulfony 1 urea. This suggests that not only the insulin secretion but also glucotoxicity plays an important role in early stage of type 1 diabetes mellitus.

The self-monitoring of blood glucose is a recommendable method for control of diabetes mellitus. But few reports of the prospective insulin sliding scale for outpatients were available in Japan. To elucidate the efficacy of the prospective insulin sliding scale for outpatients with diabetes mellitus, 14 insulin-treated patients who usedwith the prospective insulin sliding scale (scale group) and 14 insulin-treated patients who did not use with the prospective insulin sliding scale (control group) were studied over a period of 6 months.
The control group showed no significant change of Hb Aic level. But the scale group showed a significant reduction in Hb Aic level at 5 months (8.37 ± 1.14% to 7.50 ±1.42%, p<0.008). The scale group had an almost two fold increase in the frequency in hypoglycemia. However, there was no severe hypoglycamia. There were weight gains in the control group and scale group, but there was no difference between control group and scale group.
These data suggest that the prospective insulin sliding scale is useful for the better control of diabetes mellitus.

An 78-year-old woman visited our hospital with the chief complaints of abdominal pain, nausea, vomiting and diarrhea. She was operated on for gastric ulcer when she was 56 years old and complained of abdominal discomfort after operation. One day she wished to take Confrey (Symphytum officinale), but she took digitalis leaves (Digitalis purpurea) by mistake. On admission she was drowsy and the systolic blood pressure was 60 mmHg on palpation. Electrocardiograms showed bradycardia, irregular AV-nodal rhythm and ST depressions and T wave inversions with the shortening of the QT interval. Serum levels were potassium, 6.4 mEq/1 ; BUN, 34.4 mg/di ; creatinine, 1.5 mg/di ; digoxin, 2.0 ng/ml (therapeutic level, up to 2.0 ng/ml); and digitoxin, 111ng/ml (therapeutic level, up to 25 ng/ml). Hemodynamic data showed HR, 49 beats/min ; CO, 2.81/min; CI, 2.31/min/m²; SV, 57 ml/beat and SI, 47 ml/beat/m² on administration of dopamine 7μg/kg/min. So VVI pacing was started at 70 bpm and CO increased to 3.6 1/min after pacing. But unfortunately she died of ventricular fibrillation. The digitalis leaves resemble the Confrey leaves and it is easy to mix up these two plants. Although this is a rare case, digitalis intoxication can be life-threatening. So we should prepared ourselves for the treatment of patients poisoned with digitalis and other poisonous herbs.

This report concerns a 62-year-old woman with Wegener's granulomatosis. She complained cough, macroscopic hematuria and oligulia. She had a consolidation without a cavity in the left upper lung field, massive proteinuria (2.5 g/day) and massive hematuria. The serum creatinine level was 2.5 mg/dl. The C-ANCA was positive. Renal biopsy revealed crescentic glomerulonephritis. She wa treated with prednisone and trimethoprim-sulfamethoxazole. The lung shadow, proteinuria and hematuria disappeared in one month. The C-ANCA titer also decreased. The use of trimethoprimsulfamethoxazole which may eliminate an infection as a cause to promote Wegener's granulomatosis should be actively included in the conventional treatment.

We report a case of reactive hypoglycemic coma in a 77-year-old man. Seven months after partial gastrectomy for early gastric cancer, he presented with syncopal attacks and seizure. His plasma glucose and insulin levels at coma were 18 mg/dl and 27μU/ml. Insulinoma was neglected with computerized tomography, magnetic resonance imaging and angiography. Because dietary control was insufficient, oral diazoxide therapy was done. But diazoxide did not protectthe overresponse of the insulin and reactive hypoglycemia in 75 g GTT. Octreotide (100 micrograms IM) inhibited insulin release and prevented hypoglycemia. Acarbose delayed the response of insulin butdid not inhibited insulin release. However, acarbose also prevented reactive hypoglycemia. We concluded that acarbose is an effective therapy for reactive hypoglycemic coma.