This study defined the normal variation range for different subsets of T-lymphocyte cells count in two different Brazilian regions. We analysed the T-lymphocytes subpopulations (CD3+, CD4+, CD8+) in blood donors of two Brazilian cities, located in North (Belem, capital state of Para, indian background) and Northeast (Salvador, capital state od Bahia, African background) regions of Brazil. Results were compared according to gender, stress level (sleep time lower than 8 hours/day), smoking, and alcohol intake. Lymphocytes subpopulations were measured by flow cytometry. Five hundred twenty-six blood donors from two Brazilians cities participated in the study: 450 samples from Bahia and 76 samples from Pará. Most (60%) were men, 59% reported alcohol intake, 12% were smokers, and 80% slept at least 8 h/day. Donors from Bahia presented with significantly higher counts for all parameters, compared with Para. Women had higher lymphocytes levels, in both states, but only CD4+ cells count was significantly higher than men's values. Smokers had higher CD4+ counts, but sleep time had effect on lymphocytes levels only for Para's donors (higher CD3+ and CD4+ counts). That state had also, a higher proportion of donors reporting sleep time <8 h/day. The values for CD3, CD4 and CD8+ cells count were significantly higher in blood donors from Bahia than among those from Pará. Female gender, alcohol intake, stress level, and smoking were associated with higher lymphocyte counts. The use of a single reference range for normal lymphocytes count is not appropriate for a country with such diversity, like Brazil is.

Acanthosis nigricans (AN) has been recognized as a marker of insulin resistance and diabetes mellitus. We have compared frequency of race and metabolic disturbances in obese women with several degrees of AN (AN group, N = 190) to a group without AN (non-AN group, N = 61) from a mixed racial population. The groups were similar regarding age and body mass index. All patients (except the diabetic patients) underwent an oral glucose tolerance test (75 g). The racial distribution of this population was 35.1% white, 37.8% mulatto and 27.1% black and the frequency of AN was 62.5, 82.1 and 83.8%, respectively, higher in black versus white (P = 0.003) and mulatto versus white (P = 0.002) women. The frequencies of diabetes mellitus and impaired glucose tolerance were 5.8 and 12.6% in the AN group and 1.6 and 8.2% in the non-AN group, respectively (P>0.05). Fasting glucose, ß cell function determined by the homeostasis model of assessment (HOMA), fasting insulin and insulin area under the curve were similar for the AN and non-AN groups. A higher HOMA insulin resistance was observed in the AN group compared to the non-AN group (P = 0.02) and in the subgroup of highest degree of AN compared to those with other degrees. The mean lipid levels and the frequency of dyslipidemia were similar for the two groups. AN was strongly associated with the black or mulatto rather than the white race, even after taking into account the effect of age, body mass index and HOMA insulin resistance.