A cross-sectional multicentre study of randomly selected diabetic patients was performed using a standardised questionnaire and examination, to establish the prevalence of peripheral neuropathy in patients attending 118 hospital diabetes clinics in the UK. Vibration perception threshold was performed in two centres to compare with the clinical scoring systems. A total of 6487 diabetic patients were studied. 53.9% male, median age 59 years (range 18-90 years). 37.4% Type 1 (insulin-dependent) diabetes mellitus, with a median duration of diabetes 8 years (0-62 years). The overall prevalence of neuropathy was 28.5% (27.4-29.6%) (95% confidence interval) in this population. The prevalence in Type 1 diabetic patients was 22.7% (21.0-24.4%) and in Type 2 (non-insulin-dependent) diabetic patients it was 32.1% (30.6-33.6%). The prevalence of diabetic peripheral neuropathy increased with age, from 5% (3.1-6.9%) in the 20-29 year age group to 44.2% (41.1-47.3%) in the 70-79 year age group. Neuropathy was associated with duration of diabetes, and was present in 20.8% (19.1-22.5%) of patients with diabetes duration less than 5 years and in 36.8% (34.9-38.7%) of those with diabetes duration greater than 10 years. Mean vibration perception threshold measured at the great toe was 21.1 +/- 13.5 SD volts and correlated with the neuropathy disability score, r = 0.8 p < 0.001. In conclusion, diabetic peripheral neuropathy is a common complication associated with diabetes. It increases with both age and duration of diabetes, until it is present in more than 50% of Type 2 diabetic patients aged over 60 years.(ABSTRACT TRUNCATED AT 250 WORDS)

No abstract

Resonances for the ketone bodies 3-D-hydroxybutyrate, acetone and acetoacetate are readily detected in serum, plasma and urine samples from fasting and diabetic subjects by 1H n.m.r. spectroscopy at 400 MHz. Besides the simultaneous observation of metabolites, the major advantage of n.m.r. is that little or no pretreatment of samples is required. N.m.r. determinations of 3-D-hydroxybutyrate, acetoacetate, lactate, valine and alanine were compared with determinations made with conventional assays at six 2-hourly intervals after insulin withdrawal from a diabetic subject. The n.m.r. results closely paralleled those of other assays although, by n.m.r., acetoacetate levels continued to rise rather than reaching a plateau 4h after insulin withdrawal. The 3-D-hydroxybutyrate/acetoacetate ratio in urine during withdrawal gradually increased to the value observed in plasma (3.0 +/- 0.2) as determined by n.m.r. The acetoacetate/acetone ratio in urine (17 +/- 6) was much higher than in plasma (2.5 +/- 0.7). Depletion of a mobile pool of fatty acids in plasma during fasting, as seen by n.m.r., paralleled that seen during insulin withdrawal. These fatty acids were thought to be largely in chylomicrons, acylglycerols and lipoproteins, and were grossly elevated in plasma samples from a non-insulin-dependent diabetic and in cases of known hyperlipidaemia.

Single‐pulse and Hahn spin‐echo 500 MHz 1H NMR spectra of human blood plasma and isolated chylomicrons, VLDL, LDL and HDL are reported. The comparison has enabled specific assignments to be made for the resonances of individual lipoproteins in the CH2 and CH3 (fatty acid), and NMe+ 3 (phospholipid choline head group) regions of the spectra of plasma (0.8–1.3 and ∼ 3.25 ppm, respectively). Fasting, and freeze‐thawing of plasma samples led to marked changes in the intensities and linewidths of lipid resonances. Analysis of lipid resonances in the spectra of plasma in terms of individual lipoproteins may shed new light on many conditions of clinical and biochemical interest.

One- and two-dimensional (correlated shift spectroscopy) high resolution proton n.m.r. spectra of human cerebrospinal fluid (CSF) are reported. The merits of water suppression by freeze drying or irradiation, and spectral simplification by spin-echo methods, are discussed. Well-resolved resonances for a range of low molecular weight metabolites such as lactate, 3-D-hydroxybutyrate, alanine, acetate, citrate, glucose, valine and formate were observed. Resonances for glutamine were observed only from freeze dried samples. Concentrations determined by n.m.r. were in reasonable agreement with those from conventional methods. The n.m.r. spectra of CSF were related to the clinical conditions of the subjects. No resonances for citrate were present in spectra of CSF from subjects (three infants) with bacterial meningitis; high lactate and lowered glucose levels were observed. Strong resonances for glucose and glycine were observed for mildly diabetic subjects. Both the aromatic and the aliphatic regions of the CSF spectra from subjects suffering from liver failure contained distinctive features characteristic for hepatic coma: Intense resonances for lactate, alanine, valine, methionine, tyrosine, phenylalanine and histidine. In some cases guanine was also present, which does not appear to have been reported previously. The two-dimensional spectrum suggested the presence of abnormally high levels of a number of endogenous metabolites. Such assignments were not possible using one-dimensional spectra alone because of signal overlap.