Predictors of haemoconcentration at delivery: association with low birth weight.

PURPOSE: [corrected] To assess the factors associated with risk of haemoconcentration at delivery, such as initial haemoglobin levels and alterations in the HFE gene, and its effect on low birth weight in pregnant women supplemented with moderate doses of iron. METHODS: Case-control study nested in a longitudinal study conducted on 217 healthy pregnant women taking moderate iron supplementation and their newborns. Women were classified according to the risk of haemoconcentration at delivery, defined as Hb > 130 g/L. Each subject's obstetric and clinical history, smoking habit, and iron biochemical parameters (haemoglobin (Hb), serum ferritin and transferrin saturation) were recorded at 1st, 2nd and 3rd trimester and at delivery. Polymorphisms of the HFE gene (C282Y, H63D and S65C) were also measured. RESULTS: The average of iron supplementation of all the women was 43.9 mg/dia (geometric mean, 95 % CI: 43.6-44.1). Higher levels of Hb at early gestation and the presence of HFE mutations were associated with greater risk of haemoconcentration at delivery, adjusted odds ratios of 1.14 (95 % CI: 1.05-1.25) and 5.35 (95 % CI: 1.6-17.8). Haemoconcentration at delivery was associated with a greater risk of low birth weight, adjusted odd ratio of 11.48 (95 % CI: 1.13-116.6). CONCLUSIONS: Moderate daily doses of supplementary iron may be harmful for foetal growth in women with alterations in HFE gene and who started pregnancy with good haemoglobin levels. Overall, this suggests the importance of determining a woman's iron status early in her pregnancy in order to establish a more appropriate pattern of supplementation.

Response to repeated phlebotomies in patients with non-insulin-dependent diabetes mellitus.

Regardless of type, uncontrolled diabetes represents a serious disruption of fuel homeostasis with consequences throughout the body. This may hamper the applicability of predeposited autologous blood transfusion in diabetic patients because metabolic changes are expected as a consequence of repeated bleeding. We undertook this study to determine whether the presence of non-insulin-dependent diabetes mellitus (NIDDM) influences the erythropoietin (EPO) response to repeated phlebotomies with respect to normal subjects. We included 22 consecutive patients scheduled for major surgery during a 2-year period in which clinical and metabolic complications were excluded and renal and liver function was considered unaffected. Selected biochemical and hematologic variables were serially measured during donation of several units of blood in a 12- to 29-day period. Bleeding produced a significant decrease in serum glucose, cholesterol, triglyceride, and apoprotein B concentration in diabetic patients. Except for glucose, this effect was not observed in controls. Both groups were comparable with respect to initial hemoglobin concentrations and all hematologic variables measured. The decrease in hemoglobin concentration did not produce clinical symptoms in these patients, and recovery was regarded as normal in both groups. Serum EPO levels in diabetic patients were negatively influenced by the initial hemoglobin A1c (HbA1c) proportion. Moreover, three nonrespondent diabetic patients with poor glycemic control responded normally 6 to 13 months later, in a second operation, when glycemic control had improved significantly. In conclusion, NIDDM may limit the donation of requested units for major surgery only if poor glycemic control is present. When possible, phlebotomies should be delayed and metabolic control reinforced.