Association between increased levels of TNF-alpha, decreased levels of prealbumin and retinol-binding protein, and disease outcome

We determined whether there is an association between tumor necrosis factor-alpha (TNF-alpha), undernutrition [prealbumin (PA) <160 mg/L, retinol binding protein (RBP) <30 mg/L], disease stage, outcome (death or survival), and race in children with leukemia. TNF-alpha, PA, and RBP were measured in 52 patients (0.8 to 17 years old): 18 African Americans, 34 whites; 27 newly diagnosed (ND), and 25 in clinical remission (CR). Mean levels of TNF-alpha were higher in patients than in 46 healthy children (p < 0.05), but were not different between ND and CR groups. Mean acute phase proteins (APP) were different among groups: ND > CR > controls (p < 0.05). Mean levels of PA and RBP were lower in patients than in controls (p < 0.051, and tended to be higher in CR than in ND patients. African-American patients had lower concentrations of TNF-alpha, PA, and RBP but higher APP than white patients (p < 0.05). CR patients and African-American patients who died tended to have higher levels of TNF-alpha and APP, but lower PA and RBP than those who survived. A higher percentage of ND African Americans (45%) than of ND whites (13%) died. Results suggest that undernutrition and inflammation in CR patients and African Americans were associated with poor survival, and that ND African Americans have a poorer outcome than whites independently of TNF-alpha levels.

Prevalence of anaemia and iron deficiency in urban Haïtian children two to five years of age

This study was undertaken to assess the prevalence of anaemia and iron deficiency (ID) in 305 urban Haïtian children, 142 boys and 163 girls from low socioeconomic class, ranging in age from 2 to 5 years. Haemoglobin (Hb), serum ferritin (FERR), serum iron, total iron binding capacity (TIBC), transferrin saturation (TS), and red blood cell indices were measured by standard techniques. Although the means of these indices were within normal range, 58.4 per cent of children had at least one of the measurements in the abnormal range (FERR < 12 micrograms/l, TS < 12, HB < 10.7 g/l in 2 year old and < 10.9 g/dl in 3-5 year old children). The overall prevalence of anaemia (40 per cent) was slightly higher in boys (42 per cent) than in girls (36 per cent). Approximately 45 and 31 per cent of children had FERR < 12 micrograms/l TS < 12 per cent, respectively, with no difference between boys and girls. Despite the decrease in the prevalence of anaemia and ID with age, about one-third of the 5 year old children were either anaemic or iron deficient. Hypochromia and microcytosis were present in 60 and 66 per cent of children respectively. Although ID was the major cause of anaemia, protein-energy malnutrition as judged by low TIBC contributed to the high prevalence of anaemia. Megaloblastic anaemia and haemoglobinopathies did not significantly contribute to the high prevalence of anaemia. The frequency of fruit consumption, hence vitamin C, was lower in anaemic than non-anaemic children. We conclude that the eradication of anaemia and ID in this population will require improvement in overall nutritional status.

Increased circulating levels of soluble HLA class I heterodimers in patients with sickle cell disease

This study examined the presence of a persistent state of low-grade inflammation in sickle cell anemia patients by measuring circulating sHLA-I heterodimers and C-reactive protein during the steady state and after recent crises. Thirty-nine pediatric sickle hemoglobinopathy patients were studied during the steady state and 11 patients were evaluated within 1 month of a painful crisis. A disease severity score was generated for each patient, and soluble HLA-I (sHLA-I) and C-reactive protein levels were determined. Soluble HLA-I was significantly elevated in 55% of the steady-state group and in 36% of the recent-crisis group. The percentage of patients with elevated sHLA-I differed in the various disease subgroups in the steady state: 46% of Hb SS patients, 70% of Hb SC patients, 75% of Hb S beta-thal patients, and 20% of Hb SSF patients. Steady-state and recent-crisis sHLA-I levels were not significantly different. C-reactive protein levels were elevated in 11% of steady-state patients and in 9% of recent-crisis patients. Soluble HLA-I levels did not correlate with C-reactive protein levels or disease severity score, age, hemoglobin, reticulocyte count, platelet count, or white cell count. These results show that the majority of sickle hemoglobinopathy patients have elevated sHLA-I levels during the steady state and after recent crisis, suggesting the presence of chronic inflammation during the steady state.

Tumor necrosis factor alpha in children with sickle cell disease in stable condition

Tumor necrosis factor alpha (TNF-alpha) is known to induce wasting in humans and animals. This study was undertaken to determine TNF-alpha concentrations in children with sickle cell disease (SCD) and whether high TNF-alpha levels are more likely to be present in children with growth deficits, infection, or pain crisis. Tumor necrosis factor alpha was measured using enzyme immunoassay in 143 blood samples obtained from 101 children. Mean TNF-alpha levels were higher in patients (50 pg/mL) than in 21 control children (19 pg/mL) and in 26 laboratory employees (20 pg/mL). During the follow-up period, 35%, 38%, and 28% of children with SCD had infection, pain crisis, or a blood transfusion, respectively. Mean TNF-alpha concentrations were higher in children who had an infection than in those who did not. No significant effect of pain crisis or blood transfusion was observed. Tumor necrosis factor alpha concentrations were above normal (> 40 pg/mL) in 15% of controls, 34% of children with SCD, and 52% of children with SCD who had an infection and 33% of those who did not. A higher percentage of children who had elevated TNF-alpha levels had weight (46% versus 31%) or height (50% versus 28.6%) deficits than children who had normal TNF-alpha levels. These results indicate that most children with SCD in stable condition have normal TNF-alpha concentrations and that those with high TNF-alpha levels are more likely to have growth deficits.

Serum transferrin receptor concentrations in women with mild malnutrition

We determined the influence of undernutrition on blood soluble transferrin receptor (sTfR) concentrations, an indicator of iron deficiency, in 99 Zairean women (aged 16-45 y) without inflammation. They were recruited during a survey on iron deficiency in rural Bas-Zaire. sTfR was measured by enzyme immunoassay, and indicators of nutritional status [albumin, transthyretin (or prealbumin), and retinol binding protein] were measured by radial immunodiffusion. Undernutrition was diagnosed if the concentration of any one of the indicators was below normal: albumin < 35 g/L, transthyretin < 160 mg/L, and retinol binding protein < 30 mg/L. The sTfR concentration ranged from 1.89 to 19.1 mg/L (mean: 8.7 mg/L). Mean values for indicators of nutritional status, serum ferritin, and transferrin saturation were within the normal range for health subjects. Regardless of the iron status (iron sufficiency, anemia, or iron deficiency with or without anemia) and whether women were pregnant or nonpregnant, undernutrition did not significantly reduce sTfR concentrations. A higher percentage (80%) of iron-deficient women with two or three protein values below normal had sTfR concentrations > 8 mg/L (which are suggestive of iron-deficiency erythropoiesis) compared with iron-deficient women with no (72.7%) or one (66.7%) protein value below normal, anemic women (46-60%) and iron-sufficient women (18.2-36.8%). Results suggest that sTfR can be used as an indicator of iron deficiency in field studies without in-depth assessment of nutritional status. However, the effect of severe malnutrition on this index requires further investigation.

Soluble transferrin receptor as an index of iron status in Zaïrian children with malaria

This study was designed to evaluate soluble transferrin receptor (sTfR) as an index of iron status in 0.5-16-year-old Zaïrian children: 17 with symptomatic malaria, 8 with asymptomatic malaria, and 15 controls. sTfR was also measured in 20 plasma samples obtained from iron sufficient laboratory employees. sTfR, haemoglobin and ferritin were measured by enzyme-immunoassay, cyanmethaemoglobin method, and radioimmunoassay respectively. Mean haemoglobin levels were lower and ferritin higher (P < 0.001) in children with clinical symptoms of malaria than in those without malaria, and they were intermediate in those with asymptomatic malaria. Mean sTfR concentrations were similar among the three groups of children and laboratory controls. There was a considerable overlap in sTfR concentrations between the three groups of children (1.8-10.2, 2.9-11.6 and 2.97-8.95 mg 1(-1) in symptomatic malaria, asymptomatic malaria and control groups, respectively) as well as laboratory controls (1.2-7.30 mg l-1). Despite the overlap, 6 children with malaria (24%) and one control child (6.7%) had sTfR concentrations above the highest concentration found in laboratory controls. No child had serum ferritin < 12 micrograms l-1 (suggestive of iron deficiency). As expected, sTfR negatively correlated with ferritin (r = -0.230) in the overall study population of children, and with haemoglobin in children with asymptomatic malaria (r = -0.943, P < 0.05), as well as in control children (r = -0.363). All children with sTfR above normal were also anaemic. Although the upper limit of normal sTfR concentration in healthy children is unknown, using the cut-off value of adults, we conclude that sTfR might be a more sensitive index of iron deficiency than serum ferritin in patients with malaria.

Assessment of iron status of Zairean women of childbearing age by serum transferrin receptor

Soluble transferrin receptor (sTfR), previously shown to be a sensitive indicator of tissue iron deficiency, was used to assess iron status of 104 Zairean women (69 lactating, 19 pregnant, and 16 nonpregnant, nonlactating women). Thirteen iron-sufficient female laboratory employees were also studied. Mean sTfR concentrations were higher in pregnant women (9.90 mg/L) than in lactating women (8.55 mg/L), nonlactating women (7.74 mg/L), and laboratory employees (5.11 mg/L) (P < 0.005). Mean serum ferritin and transferrin saturation were lower in pregnant than nonpregnant women. sTfR negatively correlated with hemoglobin (P < 0.05) and transferrin saturation (P < 0.05), and nonsignificantly with ferritin and transferrin. With 7.26 mg sTfR/L (the upper limit of laboratory employees) as the cutoff value, sTfR identified 67% of women with tissue iron deficiency compared with 11.5-57% for transferrin saturation, hemoglobin, or ferritin. Despite the moderate sensitivity (68.5%), 90% of women with sTfR > 7.26 mg/L also had another index below normal and 54% had serum ferritin < or = 50 micrograms/L.

Reference levels of acute phase reactant proteins in healthy Zairean women in the reproductive age group

Acute phase proteins (APP) are plasma proteins secreted by the liver following inflammation. APP reference levels are unknown for healthy female Zaireans. The purpose of this study was to determine: (a) the reference levels of alpha 1-acid glycoprotein (AGP), C-reactive protein (CRP), ceruloplasmin (Cp), and haptoglobin (Hp) in healthy Zairean adult women, and (b) whether the levels are different between lactating (Group I), pregnant (Group II) and non-pregnant/non-lactating women (Group III). APP were measured by radial immunodiffusion in 180, 212 and 27 women of Groups I, II, and III respectively (age range 15-45 years). The ranges for different APP were wider than for Western adults or those obtained in plasma from normal laboratory personnel. The means and medians of AGP were lower in pregnant that non-pregnant women, and those of CRP and Cp were higher. There were no differences in mean Hp levels between the three groups. In non-pregnant women, the mean levels and 95% confidence limits of CRP, Cp and Hp were within published ranges for healthy Western adults, while those of AGP were higher. In pregnant women only AGP and Hp were within published ranges. We conclude that, at least in non-pregnant women, the same cut-off points of CRP, Cp and Hp used for Western adults may also be used in the assessment of inflammation in Zairean women.

Usefulness of serum ferritin levels in the assessment of iron status in non-pregnant Zairean women of childbearing age

We studied the usefulness of serum ferritin for the assessment of iron deficiency (ID) or ID anaemia (IDA) of 186 lactating and 27 non-lactating Zairean women (15-45 years old). Haemoglobin (Hb), serum iron (SI), total iron binding capacity (TIBC), and transferrin saturation (TS) were also measured. Participants were recruited in rural Bas-Zaire State in the summers of 1986 and 1989. Serum ferritin ranged from 10 to 360 micrograms l-1 (median 62 micrograms l-1) in lactating women and from 14.2 to 120 micrograms l-1 (median 40 micrograms l-1) in non-lactating women. While mean levels of serum ferritin and TS were within the normal range in both groups of women, those of Hb were below normal (< 12 g dl-1), partly due to inflammation. The prevalence of anaemia was 66% in lactating women and 70% in non-lactating women, and did not change with time. It was higher in women with inflammation than in those without inflammation. Although ID (ferritin < 12 micrograms l-1) was almost absent, after raising the cut-off point of ferritin to 50 micrograms l-1 in women with inflammation, it was present in 28.8% of lactating women and 52% of non-lactating women. While the prevalence of ID assessed by serum ferritin was significantly higher in lactating women studied in 1989 (40.5%) than in those studied in 1986 (13.5%), it was similar in both groups of non-lactating women. ID defined by TS < 16% was present in 41% of lactating women and 21% of non-lactating women. In the presence as well as absence of inflammation, the use of TS identified a higher percentage of lactating women with either ID or IDA than did the use of serum ferritin. We conclude that, in the studied population, unless inflammation is taken into consideration, serum ferritin has a limited value in the diagnosis of ID.

Transport proteins and acute phase reactant proteins in children with sickle cell anemia

Transport proteins, acute-phase reactant proteins (APRP), hematology, and anthropometry were studied in 34 sickle cell disease (SCD) children (20 boys, 14 girls) and 27 controls without growth deficits (13 boys, 14 girls) [corrected]. The age range was 1/2 to 16 1/2 years. Weight deficits (< 80%) by Waterlow's classification were observed in 41% of SCD boys and 25% of SCD girls, and height deficits (< 90%) were observed in 25% SCD boys and 25% girls. Mean white blood cell counts were significantly higher (P < .001) and hematocrit and hemoglobin (Hb) lower (P < .005) in SCD children than in controls. Although both groups had similar mean levels of albumin, transferrin, and APRP, SCD children had significantly lower mean levels of retinol-binding protein (RBP) (P < .001) and retinol-prealbumin (P < .001). Retinol-binding protein levels were abnormal in 18 (53%) SCD children and in only 23% controls (chi 2 = 14.06; P < 0.005); transferrin levels were abnormal in 20% of SCD children and in none of the controls. Children with SC and SF Hb phenotype had normal mean levels of RBP, whereas those with S beta thal and SS phenotype had levels below normal. Growth-retarded children by weight and height had reduced mean levels of RBP and prealbumin compared with growth-normal SCD children. The implication of primary protein-energy malnutrition on growth retardation in SCD children is under study.

Nutritional status of children with leukemia

Children with cancer represent a high-risk group for protein-energy malnutrition due to side effects associated with treatment. Assessment of nutritional status at the time of diagnosis and during treatment is, therefore, essential for planning nutritional intervention. We studied the nutritional status of 25 children with leukemia [9 newly diagnosed/relapsed (D/R) leukemic patients and 16 children with leukemia in remission (REM)]. Plasma proteins (prealbumin, PA; albumin, Alb; transferrin, Tr; retinol-binding protein, RBP) and acute phase-reactant proteins (alpha 1-acid glycoprotein, AGP; C-reactive protein, CRP; ceruloplasmin, CER) were measured by radial immunodiffusion. Results show that there were no significant deficits in anthropometric measurements among leukemic children. In contrast, the mean levels of all plasma proteins, especially PA (P < 0.005), were significantly lower in the D/R group than in the REM group. All D/R children, compared to 59% of those in remission, had PA levels < 20 mg/dl. Only the D/R group had abnormal levels of RBP, Tr, and Alb. Children who were treated with prednisone had significantly higher mean levels of PA, RBP, and AGP than those who were not receiving prednisone. The mean levels of acute phase-reactant proteins in these leukemic children were comparable to those of healthy children. We conclude that mild/moderate malnutrition is common in leukemic patients at D/R and that PA seems to be the most sensitive indicator of visceral protein status.

An epidemiological study of hemoglobin levels and prevalence of anemia in young children from Bas-Zaire

A population of 183 Zairean infants and young children (0.5 to 48 months old) who lived in rural Bas-Zaire has been examined in regard to hemoglobin (Hb), hematocrit (Hct) and mean corpuscular hemoglobin concentration (MCHC). The mean Hb was 9.03 g/dL, Hct 28.4%, and MCHC 30.7 micrograms/dL. These means are below normal by WHO and/or American Academy of Pediatrics criteria. Although there was no significant sex difference in the mean hematoglobin measurements, boys more than 24 months old had a higher mean Hb (9.98 g/dL) than girls (8.59d g/L) of the same age. Approximately 80%, 76% and 56% of children had Hb < 11 g/dL, Hct < 34% and MCHC < 32 micrograms/dL respectively. The highest prevalence of anemia was between the age of 12 and 18 months. Severe anemia (Hb < 8 g/dL) was observed in 30.65% of the study population (26.6% boys and 34.8% girls). Approximately one third of children had very low Hct levels (< or = 25%). Based on MCHC, iron deficiency was likely responsible of the anemia in 61% of the children. Both Hb (p < 0.05) and Hct positively correlated with the mother's level of education. Because of the many detrimental effects of iron deficiency in infants and young children on immunity, psychomotor, behaviour and mental development routine assessment of iron status and correction of iron deficiency should become a part of the Surveillance Programme in this population.

Impaired protein kinase C activation as one of the possible mechanisms of reduced lymphocyte proliferation in iron deficiency in mice

We investigated the effects of iron deficiency in mice on protein kinase C (PKC) activation, an enzyme required for optimal lymphocyte proliferation. C57BL/6 mice were fed either an iron-deficient diet (ID; 10 mg Fe/kg diet), a control diet (C; 50 mg/kg diet), or were pair fed (PF) to ID mice for 34 d. PKC activity was studied in spleen cells by histone phosphorylation. Iron deficiency significantly reduced cytosol activity in unstimulated cells and membrane-bound activity in cells stimulated by concanavalin A (Con A) or phorbol-12-myristate-13-acetate (PMA), and the ratio of membrane-bound over cytosol activity in mitogen-stimulated cells. In PF mice the ratio of membrane-bound activity to cytosol activity was greater than normal in Con A-treated cells and only slightly decreased in PMA-treated cells. PKC activity positively correlated with iron status. We conclude that reduced PKC activity and poor translocation results in aberrant signal transduction, which in turn might be responsible for the impaired lymphocyte proliferation associated with iron deficiency.

Iron status of Zairean pregnant women with and without serological markers of hepatitis B virus infection

We assessed the iron status of 203 Zairean pregnant women: 38 with chronic hepatitis B virus (HBV) infection (HBsAg(+)), 94 with antibodies to the surface antigen (Anti-HBs(+)) and 71 without HBV markers (HBsAg(-)/Anti-HBs(-)). Participants, age range 15-42 years and parity 1-12, were recruited from Mama Yemo Hospital in summer 1983. Haemoglobin (Hb), serum iron, total iron binding capacity and transferrin saturation (TS) were determined by standard techniques and serum ferritin (FERR) by radioimmunoassay. To rule out inflammation and/or infection which increase FERR levels, C-reactive protein (CRP) and alpha 1-acid glycoprotein (AGP) were also measured. There was no significant difference in the mean levels of any of the haematologic measurements, FERR, CRP and AGP between the three HBV groups. Women who took iron supplements had slightly higher mean levels of Hb but not serum FERR or TS than those who did not. Women with inflammation and identical HBV markers had higher mean FERR levels than those without inflammation. Neither the prevalence of anaemia, which varied between 32 and 35%, nor that of iron deficiency, which varied between 52 and 59%, differed significantly between the three groups of women. We conclude that in pregnant women, chronic asymptomatic HBV infection is not associated with a lower prevalence of iron deficiency and/or anaemia.

Influence of iron-deficiency anemia on selected thymus functions in mice: thymulin biological activity, T-cell subsets, and thymocyte proliferation

To define the effects of iron deficiency on thymulin biological activity, T-cell subsets, and thymocyte proliferation, C57BL/6 female mice at weaning were fed an iron-deficient diet (10 mg Fe/kg diet), an iron-sufficient diet (50 mg Fe/kg diet), or restricted amounts of the iron-sufficient diet (the pair-fed group) for 40 d. Iron deficiency did not reduce the concentration of either serum or intracytoplasmic thymulin. Although T-cell subsets in the thymus were not altered, both the cortical and medullar regions were depleted of thymocytes. In the spleen iron deficiency (but not underfeeding) significantly reduced the percentage of L3T4+ cells, of Lyt-2+ cells, and thus of the overall T-cell population. However, it did not affect the ratio of L3T4+ to Lyt-2+ T cells. Thymocyte proliferation was significantly reduced at the concanavalin A (Con A) dose (10 mg/L) that produced maximal stimulation in control and pair-fed mice but not at low (7.5 mg/L) or high (15 mg/L) Con A concentrations. We conclude that the impairment in immune functions associated with iron deficiency is not due to an impairment in thymic endocrine function but rather to decreased immunocompetent lymphocytes.

Macrophage function as studied by the clearance of 125I-labeled polyvinylpyrrolidone in iron-deficient and iron-replete mice

This study evaluated the effects of iron deficiency and iron repletion on in vivo macrophage function determined by the clearance of 125I-labeled polyvinylpyrrolidone (PVP). Two experiments were done. There were four groups of C57BL/6 female mice in experiment 1: the iron-deficient (ID), pair-fed (PF), control (C) and the high iron (HI) groups. In experiment 2, there were three ID groups (severe to moderate anemia), three PF, one C and four ID groups that were fed adequate iron for 14 (R-14), 7 (R-7), 3 (R-3) days before or on the day of PVP injection (R-0). The overall rate of PVP clearance from blood was lower in ID than in C or PF groups. This clearance is expressed by a constant, K, calculated from natural log (ln) of the cpm and the time postadministration of PVP that blood was drawn. The theoretical individual macrophages function (alpha PVP), derived from K and the weights of body, spleen and liver, was also lower in ID than in C or PF groups. The impairment was most severe with the most severe iron deficiency. Repletion for 7 to 15 d before PVP administration resulted in a partial correction of the clearance. Moderate undernutrition in the PF group had no effect.

Iron-binding proteins, hepatitis B virus, and mortality in the Solomon Islands

The hypothesis that serum levels of ferritin and transferrin are associated with subsequent mortality was tested in a population of Solomon Islanders who had been followed over an 8-12-year period beginning in 1966. A case-control analysis of 105 matched pairs showed that 1966-1970 levels of ferritin were higher and levels of transferrin were lower in Solomon Islanders who had died by 1978 than in matched controls who were alive in 1978. These findings support the hypothesis and, in addition, are consistent with the view that increased iron stores are associated with increased mortality. Among females, the association of ferritin with mortality was more pronounced in chronic carriers of hepatitis B virus than in noncarriers.

Impairment of blastogenic response of splenic lymphocytes from iron-deficient mice: in vivo repletion

Iron-deficiency anemia impaired the blastogenic response of splenic lymphocytes and partially purified T cells to Concanavalin A and phytohemagglutinin. The response of splenic lymphocytes and partially B cells to bacterial lipopolysaccharide was also significantly impaired. Caloric restriction in pair-fed mice did not have any significant effect. Blastogenic response to the three mitogens was restored to normal after anemic mice were fed the regular diet containing 25 to 30 mg Fe/kg (FeSO4) for approximately 10 days. We also found that in the anemic mice the mean wet weights per 100 g of body of spleen, heart, brain, and kidney increased, while those of the thymus and liver decreased. In the pair-fed mice only the mean wet weight of the liver significantly decreased. There was a small but significant decrease in the white blood count and peripheral lymphocyte count in the anemic but not the pair-fed mice. The mechanism by which iron deficiency impairs the cell-mediated immune response is discussed.