RESUMEN
AIM: Skin autofluoresence has been advocated as a quick non-invasive method of measuring tissue advanced glycosylation end products (AGE), which have been reported to correlate with cardiovascular risk in the dialysis patient. Most studies have been performed in patients from a single racial group, and we wanted to look at the reliability of skin autofluoresence measurements in a multiracial dialysis population and whether results were affected by haemodialysis. METHODS: We measured skin autofluoresence three times in both forearms of 139 haemodialysis patients pre-dialysis and 36 post-dialysis. RESULTS: One hundred and thirty-nine patients, 62.2% male, 35.3% diabetic, 59% Caucasoid, mean age 65.5 ± 15.2 years were studied. Reproducibility of measurements between the 1st and 2nd measurements was very good (r(2 ) = 0.94, P < 0.001, Bland Altman bias 0.05, confidence limits -0.02 to 0.04). However, skin autoflourescence measurements were not possible in one forearm in 8.5% Caucasoids, 25% Far Asian, 28% South Asians and 75% African or Afro Caribbean (P < 0.001). Mean skin autofluorescence in the right forearm was 3.3 ± 0.74 arbitrary units (AU) and left forearm 3.18 ± 0.82 AU pre-dialysis, and post-dialysis there was a fall in those patients dialysing with a left sided arteriovenous fistula (left forearm pre 3.85 ± 0.72 vs post 3.36 ± 0.55 AU, P = 0.012). CONCLUSION: Although skin autofluorescence is a relatively quick non-invasive method of measuring tissue AGE and measurements were reproducible, it was often not possible to obtain measurements in patients with highly pigmented skin. To exclude potential effects of arteriovenous fistulae we would suggest that measurements are made in the non-fistula forearm pre-dialysis.