Altered filtration of macromolecules due to decreased electrical charge of the glomerular basement membrane might be the initial step in the development of albuminuria in patients with Type 1 (insulin-dependent) diabetes mellitus. We therefore investigated the selectivity index, i.e. renal clearance of non-glycated plasma albumin/clearance of glycated plasma albumin in 38 patients with Type 1 diabetes mellitus. The two albumin molecules differed slightly in charge, non-enzymatic glycated albumin being more anionic at physiological pH compared with unmodified plasma albumin. Glycated albumin in plasma and urine was determined by a specific, sensitive and highly reproducible chromatographic procedure. In diabetic patients with normal urinary albumin excretion, the selectivity index was increased three-fold compared with that of non-diabetic subjects (2 p less than 0.01). A significant correlation (r = 0.53, 2 p less than 0.01) between haemoglobin A1c and selectivity index was demonstrated in these patients, indicating a change in charge-dependent renal filtration could possibly be attributed to non-enzymatic glycation of components in the glomerular basement membrane and tubuli. Diabetic patients with increased albumin excretion rate had a significantly lower selectivity index compared with patients with normal albumin excretion (2 p less than 0.01). A significant negative correlation (r = 0.85, 2 p less than 0.001, exponential curve fit) was seen between urinary albumin excretion and selectivity index in the diabetic patients, indicating that the capability of differentiating between macromolecules of different charges is again lost with increasing urinary albumin excretion.(ABSTRACT TRUNCATED AT 250 WORDS)

The transcapillary escape rate and relative plasma disappearance of glycated and non-glycated albumin were measured in 25 male Type 1 (insulin-dependent) diabetic patients using a double tracer technique. The patients were divided into three groups on the basis of their urinary albumin excretion: group 1, normal albumin excretion (less than 30 mg/24 h) (n = 8); group 2, microalbuminuria (30-300 mg/24 h) (n = 9); and group 3, clinical nephropathy (greater than 300 mg/24 h) (n = 8). Six male age-matched non-diabetic persons served as control subjects. The transcapillary escape rate of glycated albumin was similar in group 1 and control subjects (4.7 +/- 2.1 versus 5.1 +/- 1.7%), but significantly increased in group 2 (7.0 +/- 1.7%, p less than 0.05) and in group 3 (7.9 +/- 3.1%, p less than 0.05). The transcapillary escape rate of glycated albumin was slightly lower than that of non-glycated albumin in all groups, but significant only in normal control subjects. No difference in the catabolic rate of glycated and non-glycated albumin was found. We conclude that the in vivo effects of glycation on the clearance and transcapillary passage of albumin are small and not likely to play any significant role in the development of late diabetic microvascular complications.

The fractional plasma escape rates of glycated and non-glycated albumin have earlier been measured in groups of Type 1 (insulin-dependent) diabetic patients and control subjects. The escape of non-glycated albumin was similar in control subjects and normoalbuminuric patients, but elevated in patients with micro or macroalbuminuria. In all groups the escape rate of glycated albumin was lower than that of non-glycated albumin. Glycation increases the anionic charge of albumin. To assay for charge-dependent alterations of transport a selectivity index (non-glycated albumin/glycated albumin transport ratio) was determined from the disappearance data. The index was high in control subjects (1.021 +/- 0.0057 (SEM)). This reflects a mean difference between the two escape rates of 2.1% per hour (for comparison the mean of the fractional escape rate of non-glycated albumin of the normal control subjects was 4.7% per hour). The index was numerically even higher in normoalbuminuric patients (1.031 +/- 0.0047 (SEM)), but reached significantly lower levels in patients with microalbuminuria (1.013 +/- 0.0030 (SEM), p < 0.02). Patients with clinical nephropathy had very low levels indicating loss of selectivity (1.002 +/- 0.0068 (SEM), p < 0.001). This pattern accords well with measurements of renal clearance selectivity indices, suggesting a general, progressive deterioration of anionic perivascular barrier components in diabetic microangiopathy. The structural target for these changes is likely to be the glycosaminoglycans of the glomerular basal membrane and the interstitial matrix.

The frequency of nocturnal hypoglycaemia, i.e. blood glucose concentration (BG) <3.0 mmol/l, was evaluated in consecutively selected insulin‐dependent patients on multiple insulin injections (MII), n =23, or continuous subcutaneous insulin infusions (CSII), n =25. Blood was sampled hourly from 23.00 to 07.00. Seven patients (30%) on MII had at least one BG <3.0 mmol/l during the night. Eleven patients (44%) on CSII had hypoglycaemia (NS). The total number of BGs <3.0 mmol/l was higher on CSII, 42 of 225, versus 16 of 207 on MII (p<0.025). The duration of hypoglycaemia was 2 hours (range 1–6) on MII and 4 hours (range 1–7) on CSII with a maximal prevalence at 4 hours and between 5 and 7 hours, respectively (p=<0.05). The frequency of nocturnal hypoglycaemia is high in patients on intensified insulin regimens. Nocturnal hypoglycaemia occurs later in the night and is of longer duration on CSII than on MIL HbA1c, BG before bedtime and in the morning might be useful in the evaluation of nocturnal hypoglycaemia.

The selectivity index, i.e. clearance of non glycated albumin/clearance of glycated albumin was studied in fourteen patients with Type 1 (insulin-dependent) diabetes and normal urinary albumin excretion. The index was increased above one in all patients, and correlated significantly to HbA1c. It was, however, unaffected by 12 weeks of improved metabolic control with a mean decline in HbA1c of 1.9% in seven patients. We conclude that the increased electronegative charge of the glomerular filtration barrier observed in uncomplicated diabetes is related to long term metabolic control but not reversible during twelve weeks of strict metabolic control. This indicates a slow turnover of the components responsible for the increased charge selectivity in uncomplicated diabetes.

ABSTRACT. In three different diabetes centers the use of insulin at a strength of 40 IU/ml (U40) and 100 IU/ml (U100) in continuous subcutaneous insulin infusion (CSII) treatment of insulin dependent (Type I) diabetic patients was compared in a randomized cross‐over design. Forty‐six Type I diabetic patients, all previously treated with CSII for at least one year, completed consecutively two 13‐week periods of treatment with U40 and U100. Body weight and insulin requirements were identical at the end of the two periods. Slightly higher levels of glycemia were recorded during U40 when compared to U100 treatment: mean blood glucose was 142±34 (SD) vs. 133±34 mg/dl (2 p<0.01). The same tendency was observed in the mean glycosylated hemoglobin value (6.84±1.35 vs. 6.65±1.13%, 2p>0.1). There were no significant differences between U40 and U100 in the number of catheters used and the number of catheter blockages reported, while the development of subcutaneous nodules at the insertion sites was significantly more frequent during U40 treatment. It is concluded that the implementation of insulin in the strength of 100 IU/ml is as effective as the use of insulin in the strength of 40 IU/ml for CSII therapy, and might even be associated with slightly improved glycemic control and less subcutaneous side‐effects.