This study was conducted to determine whether glucose tolerance curves for parotid salivas of diabetics were different from those for non-diabetics. Since Langley, Gunthorpe, and Beall (Amer. J. Physiol., 192:482-84, 1958) have shown that the glucose threshold for parotid secretion varied directly with insulin dosage in dogs, it was thought that lack of insulin in man might so lower the parotid glucose threshold that large quantities of glucose would appear in saliva. If so, analysis of parotid fluid might be as useful as analysis of blood or urine in the diagnosis of diabetes mellitus.Early-morning samples of venous blood and parotid saliva were collected from 26 uncontrolled diabetic and 26 non-diabetic persons, aged twelve to seventy years, in the postabsorptive state just prior to, and at about 30, 60, 120, and 180 minutes after oral administration of 1 gm. * Mean and standard deviation. per kg. of body weight. At each collection time, venous blood was drawn, and approximately 5 ml. of parotid saliva stimulated with lemon juice was obtained by means of a parotid cap. Collection of saliva was completed about 7 minutes after each blood sample was drawn.Glucose concentrations were determined by the method of Saifer and Gerstenfeld (J. Lab. clin. Med., 51:448-60, 1958) on duplicate samples of 0.2 ml. of deproteinized plasma and 2 ml. of non-deproteinized parotid saliva.Glucose concentrations in parotid secretion varied directly with plasma concentrations for each subject but at a much lower order of magnitude (see the accompanying tabulation).Although average glucose concentrations in parotid saliva for diabetic subjects were slightly higher at each time period than those for non-diabetic subjects, there was considerable overlapping of curves for individuals in both groups. Therefore, it is doubtful that parotid glucose concentrations could be useful, or replace the use of plasma, in the diagnosis of diabetes mellitus.

Hypertension has recently been shown to be an important determinant of diabetic retinopathy and nephropathy. The relationship between cardiovascular risk factors and the incidence of peripheral neuropathy (PN) was examined in type 1 diabetic subjects from 27 centres participating in the EURODIAB Prospective Complications Study. PN was assessed at baseline and follow‐up using a standardised protocol involving combinations of neuropathic symptoms, tendon reflexes, age related vibration perception thresholds and autonomic function tests. Serum lipids/lipoproteins, HbA1c and albumin excretion rate (AER) were measured in a central laboratory. Of 986 subjects with no PN at baseline (mean age 31 years; mean duration 13 years), 24.6% developed PN over the follow up period (average 7.3 years). The incidence of PN was significantly positively associated with age, duration of diabetes and HbA1c at baseline. After statistical adjustment for these three factors the following baseline variables were significantly predictive of the development of PN; BMI, AER, triglyceride (p < 0.001), cholesterol and systolic BP (p < 0.01). This prospective study shows that over a 7‐year period, about one quarter of type 1 diabetic patients will develop peripheral neuropathy; age, duration of diabetes and poor glycaemic control being major determinants. The development of PN is also associated with potentially modifiable cardiovascular risk factors such as serum lipids, BP, BMI and AER supporting risk reduction strategies in its prevention. Furthermore, these findings support the role of vascular factors in pathogenesis of PN.

The relationship between cardiovascular risk factors and the incidence of peripheral neuropathy (PN) was examined in type 1 diabetic subjects from 27 centres participating in the EURODIAB Prospective Complications Study. PN was assessed at baseline and follow-up using a standardised protocol involving combinations of neuropathic symptoms, absent tendon reflexes, age related vibration perception thresholds (VPT) and autonomic function tests. Serum lipidsllipoproteins, HbA,, and albumin excretion rate (AER) were measured in a central laboratory. Of 1195 subjects with no PN at baseline (mean age 30.6 years; mean duration 12.4 years), 24.5% developed PN over the follow up period (average 7.3 years). In those with no baseline abnormalities 24.4% developed neuropathic symptoms, 19.5% had absent reflexes, 21.2% had abnormal VPT and 15.5% abnormal autonomic function at follow up. The incidence of PN was significantly positively associated with age, duration of diabetes and HbA,, at baseline. After statistical adjustment for these three factors the following baseline variables were significantly predictive of the development of PN; BMI, AER, triglyceride (p