For the past 10-15 years all the children at our unit with insulin-dependent diabetes mellitus have been repeatedly followed-up with renal function tests. Renal biopsy, examined by light and electron microscopy, was included in the follow-up of 36 adolescents and young adults, aged 13-25 years, with a disease duration of 7-19 years. All subjects had undergone at least three renal function tests before biopsy and none had persistent microalbuminuria. Renal function was evaluated as glomerular filtration rate and effective renal plasma flow determined by clearances of inulin and para-amino hippuric acid. Glomerular filtration rate and filtration fraction were increased before and at the time of the biopsy. Glomerular basement membrane thickness (331-858 nm) and mesangial matrix volume fraction (7.4-17.1%) were increased. Long-term hyperfiltration and hyperperfusion before biopsy correlated inversely with mean glomerular volume. Increased filtration fraction before the biopsy correlated directly with mean of all HbA1c (r = 0.485, p< 0.01) and both variables correlated directly with mesangial matrix volume fraction, basement membrane thickness and structural index (r = 0.433, p < 0.01 and r = 0.626, p < 0.001, respectively). Urinary albumin excretion rate correlated directly with foot process width (r = 0.645, p < 0.001). By multiple regression analysis the most important variable for the increase in basal membrane thickness was the metabolic control while the mean of previous filtration fraction was most important for the increase in mesangial matrix volume. In conclusion, although none of the patients showed constant microalbuminuria, early diabetic structural changes were evident with basal membrane thickening and increased mesangial matrix volume. The structural changes related to long-standing hyperfiltration and poor metabolic control.

The pattern of fall in B-cell function measured as plasma and 24 h urinary C-peptide excretion, as well as levels of islet cell antibodies, insulin antibodies and metabolic parameters, were followed for two years in 39 children aged 1-17 years prospectively from clinical onset of Type 1 (insulin-dependent) diabetes. At onset 32/36 patients had measurable plasma C-peptide (median 0.13 nmol/l). Maximum values of fasting and postprandial plasma C-peptide were reached at a median duration of three months. Thereafter both plasma and urinary C-peptide declined linearly. The median value of the rate of fall in postprandial plasma C-peptide was 0.019 nmol.1-1.month-1. Age at onset was positively correlated to the maximum value of postprandial plasma C-peptide in each patient (rs = 0.57, p = 0.0001) and throughout the observation time positively correlated to fasting and postprandial C-peptide and to the 24 h urinary C-peptide excretion (rs range 0.35-0.70, p = 0.03-0.0001). The rate of fall of postprandial C-peptide was unrelated to age at onset and was strikingly parallel in different age groups. Islet cell antibodies were present in 87% of the patients at onset and decreased to 38% at 24 months. Islet cell antibody titres were not correlated to age at onset or to plasma or urinary C-peptide at any single observation. However, islet cell antibody negative patients had significantly higher (p less than 0.05) postprandial plasma C-peptide values at 1, 9, and 12 months of duration, compared to islet cell antibody positive patients.(ABSTRACT TRUNCATED AT 250 WORDS)

This is a prospective study of the incidence of insulin-dependent diabetes mellitus (IDDM) in children 0-14 years of age, including all newly diagnosed cases in the whole of Sweden from July 1, 1977 until June 30, 1980. All 45 Swedish departments of paediatrics participated. During the three-year-period studied, 1108 Swedish children, 0-14 years of age had their onset of diabetes. That means around 369 new diabetics yearly in the age groups studied. The mean yearly incidences in the years 1977-80 were 22.6, 22.8 and 22.6 per 100000 children, respectively. Mean prevalence on June 30, 1980 and 1.48 per 1000 children 0-14 years with a wide range of 0.71-2.65. The age distribution at onset showed a gradual increase and peak incidences at 11 years of age for the girls and 4 and 13 years of age for the boys. There was a consistently higher incidence for boys in the younger age groups during the three-year-period studied. Peak incidences of new cases were reached in January, March and July through October for the age groups 5-9 and 10-14 years of age. No such seasonal variation was seen for children 0-4 years of age. The cumulative incidence of IDDM at 14 years of age was 3.2 per 1000 for the boys and 2.9 per 1000 for the girls. The degree of ascertainment in this study was 93.4%.

In order to study the long-term development of diabetic neuropathy in children with newly diagnosed diabetes mellitus, 144 children were entered in a prospective study of nerve conduction and autonomic nervous function. Neurophysiological recordings of nerve conduction and parasympathetic function (R-R variations) were made at onset of diabetes and after 2, 5 and 10 years. Low sensory nerve conduction and autonomic dysfunction were found in approximately 25% of the children at onset of diabetes when the patients were not yet in complete remission. During years 0-2, an initial improvement of sensory conduction velocities was found. After 2 years, deteriorations in sensory and motor nerve conduction and autonomic nerve function were common and further deterioration was seen over time. A correlation was found between nerve conduction and glycaemic control.

We report a retrospective study of diabetic children, 0--14 years of age, from seven Swedish departments of paediatrics. There were 359 new cases in the years 1970--1975. Notification suggested that there was a mean yearly incidence of 19.6 cases per 100 000 with a year to year variation of 10.0--26.4 per 100 000. Consequently about 330 new cases of childhood diabetes would be expected in Sweden every year. Incidence varied considerably between different geographical areas. The age distribution was bimodal with a main peak at about 12 years and another peak at about 7 years. There was some evidence for clustering of new cases in January and the autumn. The mean prevalence of childhood diabetes in the seven districts was 1.3 per 1 000.