Acta Pædiatrica ISSN 0803–5253 REGULAR ARTICLE Diabetes induced immunological and biochemical changes in human colostrum G Morceli1, EL França2, VB Magalhães1, DC Damasceno1, IMP Calderon1, AC Honorio-França (denifran@terra.com.br)2 1.Post Graduate Program in Gynecology, Obstetrics and Mastology of Botucatu Medical School, Sao Paulo State University ⁄ Unesp, Sao Paulo, Brazil 2.Institute of Biological and Health Science, Federal University of Mato Grosso, Pontal do Araguaia, Mato Grosso, Brazil Keywords Antibody, Colostrum, Complement protein, Diabetes, Enzymes Correspondence AC Honorio-França, Instituto de Ciências Biológicas e da Saúde, UFMT, Pontal do Araguaia MT, Rodovia MT100, Km 3,5 s ⁄ no, Pontal do Araguaia, Mato Grosso, Brazil. Tel: 55-66340121121 | Fax: 55-6634021117 | Email: denifran@terra.com.br Received 16 August 2010; revised 14 September 2010; accepted 15 October 2010. DOI:10.1111/j.1651-2227.2010.02070.x ABSTRACT Aim: This article describes the changes and relationships between biochemical and immunological parameters in the colostrum and serum of diabetic women. Methods: Colostrum and blood samples were collected from 30 diabetic and 15 normoglycaemic women. Glucose, total protein, antibody, complement proteins (C3 and C4), fat and calorie content, amylase, lipase and superoxide dismutase (SOD) were determined. Results: Glucose was higher in both the colostrum and serum of diabetic mothers compared to that of their normoglycaemic counterparts. In both groups, total protein was higher in colostrum than in serum. IgA and IgG were lower in the colostrum of hyperglycaemic mothers, whereas IgM did not vary between the groups. Colostral C3 protein was significantly lower in diabetic mothers, but colostral C4 protein was similar between normoglycaemic and hyperglycaemic mothers. Fat content was lower in the colostrum of the diabetic mothers, although calorie content did not vary between the groups. Amylase was lower in colostrum than in serum in both groups. Lipase was higher in the colostrum and serum of diabetic mothers. Colostral SOD was similar between the groups. Conclusions: Our results support the hypothesis that the colostrum of diabetic mothers suffers biochemical and immunological alterations that affect the levels of its components. INTRODUCTION Breastfeeding promotion is an important public health strategy to mitigate infant and child morbidity and mortality as well as maternal morbidity, thereby averting healthcare costs (1). Besides being the safest and most natural way to feed a neonate (2), breastfeeding has been proven to protect neonate against a wide range of infectious and noninfectious diseases. Therefore, efforts have been directed to identify the various immunoactive substances in human breast milk that account for the protective effects observed (2,3). Human milk is important to nutrition and immunological defence of human infants. It has an appropriate balance of nutrients provided in easily digestible and bioavailable forms (4). Even colostrum, the first lacteal secretion produced, contains the nutrients needed for growing infants Abbreviations ANOVA, Variance analysis; C3, Complement 3; C4, Complement 4; CuZn-SOD, CuZn-superoxide dismutase; EDTA, Ethylenediamine tetraacetic acid; GI, Glycaemic index; GTT, Glycaemia tolerance test; IgA, Immunoglobulin A; IgG, Immunoglobulin G; IgM, Immunoglobulin M; NBT, Nitroblue tetrazolium; SD, Standard derivation; sIgA, Secretory immunoglobulin A; SOD, Superoxide dismutase. 550 and provides adequate amounts of lipids, carbohydrates and proteins (1). Antibodies that are specific against microorganisms such as viruses, bacteria and parasites are also found in human milk and may colonize newborn gut just after delivery. Human milk is particularly rich in secretory IgA antibodies (sIgA) (1,2,5), which play a protective role against several microorganisms (6,7). It also contains other immunoglobulin isotypes, such as IgG and IgM, which provide complementary protection in the respiratory mucosa and gastrointestinal tract of newborns (8). Colostrum and human milk also contain complement system proteins, particularly C3 and C4. Although found at lower amounts in milk than in blood, these proteins act primarily as opsonins (9). They are thus important for phagocytosis and microbicidal activity (6). Other studies have shown that, under specific circumstances, IgA, complement factors and receptors participate in the IgA-mediated triggering of the effector function of cells (10). Breastfeeding reduces the incidence of acute illnesses and likely decreases the risk of a number of chronic diseases. Earlier studies on the long-term effects of breastfeeding showed that it affects blood pressure, obesity ⁄ overweight and diabetes (11). Clinical, epidemiological and experimental studies suggest that the baby’s diet is implicated in the ª2010 The Author(s)/Acta Pædiatrica ª2010 Foundation Acta Pædiatrica 2011 100, pp. 550–556 Diabetes and changes human colostrum Morceli et al. etiopathogenesis of diabetes (12). After adjusted for birth weight, parental diabetes, socioeconomic status and body size, breastfeeding is in fact associated with decreased risk of type 2 diabetes later in life (11). In contrast to breastfed infants, formula-fed infants have higher concentrations of blood glucose (12). Diabetes mellitus is a metabolic disease characterized by elevated blood glucose levels. It results from the absence or inadequate pancreatic insulin secretion with or without concurrent impairment of insulin action (1). Despite the importance and high incidence of this disorder, the immunoreactive proteins, carbohydrate, fat and enzymes in the colostrum of diabetic women are still not understood. In this article, we describe changes and relationships among carbohydrate, immunoreactive protein, enzyme, fat and calorie levels in the colostrum of normoglycaemic and hyperglycaemic women. Serum samples of the subjects were also studied to allow comparisons between the levels of the components studied. MATERIALS AND METHOD This cross-sectional study evaluated 45 diabetic and normoglycaemic women (18–35 years of age) treated at the Diabetes and Pregnancy Service of the Obstetrics Discipline of Botucatu Medical School, UNESP, Botucatu, SP. The volunteers signed an informed consent form before entering the study, which was approved by the local ethics committee. All the women were evaluated at the service, and those with hyperglycaemia were included in a treatment protocol to control maternal glycaemia (13). Subject evaluation During prenatal care, 30 pregnant volunteers were diagnosed with diabetes, and 15 were confirmed to be normoglycaemic by the Glycaemia Tolerance Test (g-OGTT 100 g) and by their glycaemic profile (GP) (13). Colostrum and blood samples from these subjects were analyzed according to maternal glycaemic status: normoglycaemic (normal 100 g-OGTT and normal GP; n = 15) and diabetes clinical (abnormal pre-pregnancy 100 g-OGTT and insulin dependent; n = 30). The mean and standard deviation for gestational age were 38 ± 0.8 weeks in normoglycaemic and 37 ± 0.2 weeks in hyperglycaemic groups and newborns’ birth weight were 2048.3 ± 411.8 in normoglycaemic and the 3140.1 ± 492.1 in hyperglycaemic groups. The subjects continued attending the facility, irrespective of diagnosis, and the hyperglycaemic patients followed a specific treatment for glycaemic control (13). The variables controlled in both groups during pregnancy were smoking status (yes ⁄ no), arterial hypertension (yes ⁄ no) and glycaemic index (GI), which was the mean level of plasma glucose measured over the gestation. GI was classified as adequate (GI < 120 mg ⁄ dL) and inadequate (GI ‡ 120 mg ⁄ dL) (13). Colostrum sampling About 48–72 h post-partum, 15-mL colostrum was collected from the volunteers. Supernatant was obtained by colostrum centrifugation at 160 G for 10 min at 4C. The upper fat layer was discarded, and the aqueous supernatant was stored at )80C for later biochemical and immunological analyses. Blood sampling We collected 15 mL of blood from each mother in tubes without anticoagulant. We centrifuged the blood samples at 160 G for 15 min, until serum separation. Serum samples were stored individually at )80C for further glucose, enzyme and protein determination. Glucose determination Glucose levels were determined by the enzymatic system. Samples of 20 lL colostrum ⁄ serum, standard of 100 mg ⁄ dL (Doles), were placed in 2.0 mL phosphate buffer solution (0.05 M, pH7.45, with aminoantipyrine 0.03 mM, 15 mM sodium p-hydroxybenzoate, 12 kU ⁄ L glucose oxidase and 0.8 kU ⁄ L peroxidase). The suspensions were mixed and incubated for 5 min at 37C. The reactions were read on a spectrophotometer at 510 nm. Total protein determination Total protein was determined by the colorimetric method. Samples of 20 lL of colostrum ⁄ serum, standard of 4 g ⁄ dL (laboratory test), were placed in 1.0 mL Biuret reagent (ions of copper in alkaline medium). The suspensions were mixed and incubated for 10 min at 37C. The reactions were read on a spectrophotometer at 545 nm. Separation of colostrum and blood leucocytes The colostrum and blood leucocytes were separated by a Ficoll-Paque gradient (Pharmacia, Upsala, Sweden) as described by Honorio-França et al. (6). IgA, IgM and IgG determination Serum ⁄ colostrum concentrations of IgA, IgG and IgM were determined by quantitative radial immunodiffusion (RID), according to Mancini et al.’s technique (14). A tube containing 10 mL of 1% agarose was heated to fusion in a double boiler and then transferred to bath at 56C for temperature stabilization. Anti-human IgA, lamb serum (Biolab), antihuman IgM (Sigma) and anti-human IgG (Sigma) antibodies were added to the agarose and mixed by tube inversion. The mixture was placed between two glass plates separated by a spacer. After solidification, the plates were perforated and the samples applied. Antibody content in the colostrum samples was determined using the Kallestad standard curve. C3 and C4 determination The concentrations of C3 and C4 were determined by the turbimetric method. Colostrum and serum samples were diluted at 1:12 (v ⁄ v) with a saline solution (9 g ⁄ L). C3 and C4 levels were determined in sample supernatants using C3 and C4 antisera (Bioclin) diluted at 1:12 (v ⁄ v). A calibration curve obtained by the Multical (Bioclin) calibrator was used to determine the standard curve. Samples of 10 lL of colostrum standard and positive or negative control sera (Bioclin) were placed in 500 lL buffer solution (sodium ª2010 The Author(s)/Acta Pædiatrica ª2010 Foundation Acta Pædiatrica 2011 100, pp. 550–556 551 Diabetes and changes human colostrum Morceli et al. chloride 0:15 mol + L, Tris 50 mmol + L, 6.0000 PEG 50 g + L, sodium azide 15:38 nM). The suspensions were mixed and incubated for 15 min at 37C. The reactions were read on a spectrophotometer at 340 nm. Creamatocrit analysis Colostrum samples were water-bath-heated at 40C for 15 min and subjected to vortex mixing. Capillary tubes (2 lL) were filled to approximately three quarters with the samples, sealed with sealing wax and centrifuged for 15 min. Centrifugation separated the samples into cream and serum. The cream column and the total column were measured, and fat and Kcal content calculated using the following formulae: Fat content ¼ % cream  0:59=1:46; where the % cream = cream column (mm) · 100 ⁄ total column (mm); Kcal=L ¼ ð66:8%  % creamÞ þ 290: Amylase determination Amylase was determined by the colorimetric method. A volume of 500 lL substrate (amide 0.4 g ⁄ L + phosphate buffer 100 nM, pH 7.0) was incubated for 2 min at 37C. Samples of 10 lL colostrum ⁄ serum were inoculated in the substrate and incubated for 7 min and 30 s at 37C. After this period, 500 lL of iodine solution was added (50 mM). The suspensions were mixed and read on a spectrophotometer at 660 nm. A control assay was carried out using only the substrate and iodine. Lipase determination Lipase was determined by the colorimetric method. Samples of 1.0 mL colostrum ⁄ serum were mixed with Tris buffer (100 mM hydroxymethylamine methane, pH 8.5), phenylmethyl sulfonyl fluoride (8 mM) and DTNB (3 mM dithionitrobenzoic acid, 100 mM sodium acetate, pH 6.0) and incubated for 2 min at 37C. After this period, 100 mL of tributyrate pyridine propanol (20 mM surfactant) was added to the solution and stirred for 30 min at 37C. The suspensions were subsequently added with 2.0 mL of acetone (p.a.), homogenized and kept still at room temperature for 3 min. The suspensions were then centrifuged at 400 G for 5 min and read at 410 nm. A control assay was carried out without phenylmethyl sulfonyl fluoride (enzymatic inhibitor). CuZn-superoxide dismutase determination (CuZn-SOD – E.C.1.15.1.1) Analysis of the CuZn-SOD enzyme was performed using the nitroblue tetrazolium (NBT) reduction method (Sigma). Spectrophotometric reading was performed at 560 nm (15). The individual samples were placed in glass tubes, with another tube containing a standard solution. Each tube contained 0.5 mL of the sample, and the standard tube contained 0.5 mL of hydro-alcoholic solution. Next, 0.5 mL of chloroform-ethanol solution (1:1 ratio) and 0.5 mL of reactive mixture (NBT increased by EDTA) were added to the 552 tubes. The experimental and standard solutions received 2.0 mL of buffer carbonate, and the pH was increased to 10.2 after the addition of hydroxylamine. The tubes remained still at room temperature for 15 min and were subsequently read at 560 nm. Superoxide dismutase (SOD) was calculated as follows: CuZn-SOD = (Ab standard – Ab sample ⁄ Ab standard) · 100 = % reduction of NBT ⁄ CuZn – SOD. The results were expressed in international units (IU) of CuZn-SOD. Statistical analysis Two-way ANOVA was used to evaluate glucose, antibody concentration, complement protein, calories, fat, amylase, lipase and SOD, considering the glycaemia level as one factor and the biological materials (colostrum or serum) as the other. Statistical significance was considered when p < 0.05. P-values of ‘p = 0.0000’ were recorded as ‘p < 0.0005’. RESULTS Glucose concentration This was higher in both the colostrum and serum of diabetic mothers than in the respective samples from normoglycaemic mothers (Table 1). Total protein concentration This was similar between normoglycaemic and diabetic mothers. Irrespective of glycaemia, total protein levels were higher in colostrum than in serum (Table 1). The effect of hyperglycaemia on colostrum and blood leucocytes Retrieval and viability of colostrum leucocytes were not affected in diabetic group compared to normoglycaemic (p < 0.05 – Table 1). Immunoglobulin concentration Diabetic mothers had lower levels of IgA and IgG in colostrum and IgG in the serum. Serum IgA did not vary between the groups. IgA was predominantly found in colostrum, whereas IgG was mostly found in serum (Table 2). IgM levels in colostrum and serum were similar between the groups. Serum IgM levels were lower in diabetic mothers than in normoglycaemic mothers (Table 2). Complement concentration In colostrum, the levels of C3 protein were significantly lower for diabetic mothers than for normoglycaemic mothers, but in serum samples this was similar between the groups. Only in normoglycaemic mothers were C3 levels higher in colostrum than in serum. C4 protein in colostrum and serum did not vary between normoglycaemic and diabetic mothers (Table 2). Fat and calorie content As shown in Figure 1, fat concentration was lower in the colostrum of diabetic mothers, but calorie level did not vary between the groups. ª2010 The Author(s)/Acta Pædiatrica ª2010 Foundation Acta Pædiatrica 2011 100, pp. 550–556 Diabetes and changes human colostrum Morceli et al. Table 1 Glucose level, total protein concentration and count and viability of leucocytes in colostrum and serum from normoglycaemic and hyperglycaemic mothers Parameter Glucose level (mg ⁄ dL) Total protein (g ⁄ dl) Leucocytes count (·106 cells ⁄ ml) Viability of Leucocytes (%) Colostrum Serum Colostrum Serum Colostrum Serum Colostrum Serum Normoglycaemic Hyperglycaemic Statistical 69 76 15.4 6.5 4.9 5.1 96 97 147 ± 35.8* 122 ± 12.7* 13.3 ± 0.9 5.7 ± 0.8† 4.8 ± 0.6 5.2 ± 0.6 93 ± 3.2 95 ± 4.1 F F F F F F F F ± ± ± ± ± ± ± ± 22.0 5.4 1.8 0.5† 0.5 0.7 4.7 5.4 = = = = = = = = 0.52; p = 0.51 (comparing the colostrum and serum) 16.7; p = 0.0008 (comparing the groups) 1.066; p = 0.3152 (comparing the colostrum and serum) 42.6561; p = 0.0000 (comparing the groups) 0.55; p = 0.058 (comparing the colostrum and serum) 1.718; p = 0.320 (comparing the groups) 1.85; p = 0.71 (comparing the colostrum and serum) 0.8424; p = 0.9160 (comparing the groups) Data presented as mean ± standard deviation (SD). *Statistically significant differences between the normoglycaemic and hyperglycaemic mothers, considering the same sample (colostrum or serum). † Comparing the colostrum and serum, considering the same group (normoglycaemic and hyperglycaemic mothers). Table 2 Immunoglobulins and complement protein concentrations in colostrum and serum from normoglycaemic hyperglycaemic mothers Parameter IgA (mg ⁄ dL) IgG (mg ⁄ dL) IgM (mg ⁄ dL) C3 (mg ⁄ dL) C4 (mg ⁄ dL) Colostrum Serum Colostrum Serum Colostrum Serum Colostrum Serum Colostrum Serum Normoglycaemic Hyperglycaemic Statistical 397.8 55.4 169.8 2014.6 36.2 34.5 164.2 134.5 121.6 130.0 298.6 ± 33.9* 46.2 ± 4.3† 71.0 ± 20.6* 1585.5 ± 213.5† 39.1 ± 7.9 35.7 ± 4.6† 142.5 ± 11.5 135.7 ± 14.6 127.3 ± 10.9 116.6 ± 27.0† F F F F F F F F F F ± ± ± ± ± ± ± ± ± ± 52.0 7.1† 67.3 161.3† 8.2 6.5 11.0 6.5† 11.9 1 2.2 = = = = = = = = = = 378.638; p = 0.0000 (comparing the colostrum and serum) 10.0067; p = 0.0004 (comparing the groups) 168.587; p = 0.00005 (comparing the colostrum and serum) 6.3690; p = 0.0036 (comparing the groups) 9.5539; p = 0.0035 (comparing the colostrum and serum) 1.6181; p = 0.2061 (comparing the groups) 7.8195; p = 0.0071 (comparing the colostrum and serum) 0.7424; p = 0.5150 (comparing the groups) 23.1283; p = 0.0001 (comparing the colostrum and serum) 0.6473; p = 0.5326 (comparing the groups) Data presented as mean ± standard deviation (SD). *Statistically significant differences between the normoglycaemic and hyperglycaemic mothers, considering the same sample (colostrum or blood). † Statistically significant differences between colostrum and serum, considering the same group (normoglycaemic and hyperglycaemic mothers). Amylase concentration Both groups had similar amylase concentration, generally lower in colostrum than in serum (Table 3). Lipase concentration Within each group, lipase levels were similar between serum and colostrum, but it was usually higher in diabetic mothers (Table 3). CuZn-SOD concentration In colostrum samples, CuZn-SOD level was similar between the groups, but in serum it was higher in the diabetic group than in the normoglycaemic one. Only normoglycaemic mothers had higher CuZn-SOD levels in colostrum than in serum (Table 3). DISCUSSION Other studies have attempted to elucidate the effects of lactation on maternal glucose metabolism (12). Along these lines, we demonstrate the effects of mother’s hyperglycaemia on the biochemical and immunological composition of colostrum. Similar to the milk of diabetic women (16), glucose concentration in the colostrum of diabetic mothers contained increased glucose levels. Protein concentration in colostrum was not affected by diabetes, and in both groups it was at higher levels than in serum. Earlier studies on colostrum macronutrients report that in normoglycaemic women this secretion contains low protein concentration, whereas in diabetic women, the protein levels showed within the reference limits (17). In fact, protein and glucose levels in the colostrum of diabetic women are likely maintained at normal levels and at constant ratio with capillary concentrations (18). This suggests that adequate glycaemic control can correct any abnormalities in milk composition (18). The improper control of glycaemia may result in undesirable consequences such as compromised breastfeeding because of delayed lactogenesis transition from phase I to II (19). There is an ever-increasing interest in understanding the breastfeeding effects on offspring as well as the mechanisms involved. A number of studies show the role of breastfeeding in growth promotion and against infections and diseases such as diabetes (12). Colostrum and human milk have been thoroughly studied in recent years, ever since their action against infections was confirmed. Human milk was first used clinically as a vehicle for passive immunity transfer, but its immune components are now known to be highly immunoreactive, exhibiting time-dependent alterations (20). ª2010 The Author(s)/Acta Pædiatrica ª2010 Foundation Acta Pædiatrica 2011 100, pp. 550–556 553 Diabetes and changes human colostrum 6 Morceli et al. cells (20), neutralize toxins, prevent attack and penetration by viral infections (21), in addition to avoiding tissue damage and energy loss (2). These functions are complemented by IgM and IgG antibodies (9) and, in particular, the complement proteins C3 and C4 (9). Because changes in blood antibody levels may be related to changes in B lymphocytes (22), the low IgG and IgM concentrations in the serum of diabetic mothers suggest that hyperglycaemia could affect this lymphocyte class. The complexity of human milk makes this secretion the ideal food source for babies for at least their first 6 months of life. This early nutrition is an important environmental input that can induce lifelong effects on metabolism, growth and major disease processes, such as diabetes mellitus (1). The amount and composition of milk is probably independent of the mother’s diet. Milk composition changes during lactogenesis, and these changes can be used as biochemical markers of the onset of milk secretion (22). Normal baby development is sustained by the balanced gain of fat and calories provided by breastfeeding. Even so, fat content is lower in the colostrum of diabetic mothers, as shown in the present and in other studies (17). Conversely, calorie concentration did not vary between normoglycaemic and hyperglycaemic mothers. This is possibly a result of the higher lipase levels in the colostrum of diabetic women than in their normoglycaemic counterparts. It has been suggested that high lipase levels compromise etherification and fatty acid synthesis in the mammary gland, thereby altering milk composition (17). Our results corroborate the hypothesis that diabetes changes lipid metabolism in the mammary gland. Enzymatic action for converting fat into calories is likely accelerated, thereby ensuring energy intake for growth and proper development of newborns of diabetic mothers. Human colostrum is also rich in biologically active molecules, which are essential for antioxidant functions. Their soluble components act in a child’s gut without, however, provoking an inflammatory response (23). These compounds may be enzymes that are important as immune protectors (2) or for infant development. This is the case for amylase, which increased activity in pregnant diabetic women (24). We did not detect differences in amylase concentration in the colostrum of normoglyacemic and hyperglycaemic mothers, but these levels were lower than in human serum. A 5 Fat (%) 4 * 3 2 1 0 Normoglycaemic Hyperglycaemic Colostrum 800 B Calorie (Kcal) 600 400 200 0 Normoglycaemic Hyperglycaemic Colostrum Figure 1 Fat (A) and calories (B) in the colostrum of normoglycaemic and hyperglycaemic mothers. Data presented as mean ± standard deviation (SD). *p = Statistically significant differences between control and hyperglycaemic groups. F (A) = 1.4341; p = 0.2499; F (B) = 6.1746; p = 0.0249. The immunoreactive proteins (IgA, IgG and C3) in colostrum may be lowered by diabetes, as we observed in the study. This finding disagrees with other studies that found secretory IgA concentrations in the colostrum of insulindependent diabetic women comparable with those measured in the colostrum of normoglycaemic women, neither of which are distinguishable from serum IgA levels in a reference population (16). Human milk is particularly rich in secretory IgA (5), which can block bacterial adherence to human epithelial Table 3 Amylase, lipase and superoxide dismutase (CuZn-SOD) concentrations in colostrum and serum from normoglycaemic and hyperglycaemic mothers Enzymes Amylase (U ⁄ dL) Lipase (UI) CuZn-SOD (U ⁄ mg protein) Colostrum Serum Colostrum Serum Colostrum Serum Normoglycaemic Hyperglycaemic Statistical 54.1 73.1 5.8 4.8 53.6 35.7 55.5 72.4 8.5 9.3 59.1 61.7 F F F F F F ± ± ± ± ± ± 9.4 3.3* 1.5 1.9 7.1 9.1* ± ± ± ± ± ± 9.5 6.5* 1.3† 4.3† 1.2 11.4† = = = = = = 20.0515; p = 0.0004 (comparing the colostrum and serum) 0.0152; p = 0.8987 (comparing the groups) 0.0015; p = 0.9687 (comparing the colostrum and serum) 4.2307; p = 0.04504 (comparing the groups) 11.49; p = 0.02348 (comparing the colostrum and serum) 4.8629; p = 0.01028 (comparing the groups) Data presented as mean ± standard deviation (SD). *Statistically significant differences between colostrum and serum, considering the same group. †Statistically significant differences between the normoglycaemic and hyperglycaemic mothers, considering the same sample (colostrum or blood). 554 ª2010 The Author(s)/Acta Pædiatrica ª2010 Foundation Acta Pædiatrica 2011 100, pp. 550–556 Diabetes and changes human colostrum Morceli et al. Amylase activity is high in colostrum and may confer to breastfed infants the ability to digest starch (24,25). This is therefore indispensable to their normal growth and development (25). In addition, some studies show that amylase associated with the SOD enzyme exerts a protective action against pancreatic lesions (15). In the present study, we demonstrated that SOD concentration in colostrum did not vary between normoglycaemic and hyperglycaemic mothers but that it was higher than in human serum. The activity of SOD in human milk appears to be 10–25 times higher than in serum (23) and may change significantly during lactation to meet the different needs of newborn development (26). Other studies show that the highest activity of this enzyme in milk occurs at 3 weeks of lactation and that it decreases after 4 months of lactation (26). Diabetic patients usually present delayed defence factors, decreased antioxidant defence enzymes as well as failure on oxidative burst, an important biochemical and immunology pathway. Hyperglycaemia severely compromises the different endogenous antioxidant defences of diabetic patients with diabetes (27). These defences may involve enzymatic pathways (27), including those of CuZn-SOD. Changes in SOD activity have been found in chemically induced diabetic animals (28). However, the effects of diabetes on SOD synthesis are a matter of controversy. Some authors reported increased levels of SOD, whereas others observed a decrease or even normal SOD concentrations in diabetes (29). Nevertheless, the antioxidant capacity of the colostrum of diabetic mothers, which increases in mature milk, is vital for the newborn in the first days of life (30). Our findings support the hypothesis that the production of milk components changes in diabetic mothers because of alterations in glucose metabolism. Therefore, adequate maternal glycaemic control of diabetic mothers is crucial to ensure that the nutritional needs of newborn babies are met and that the immunity components are properly provided. Despite the abnormalities in biochemical and immunological components, women with diabetes should be strongly encouraged to breastfeed their children. Besides being an excellent food source for newborns, breast milk decreases the high rates of maternal and infant complications. In addition, the rate of growth of breastfed infants is related to the total amount of milk they consume rather than the concentration of fat, proteins or carbohydrates in the milk. ACKNOWLEDGEMENTS We are very grateful to the Diabetes and Pregnancy Service, Obstetrics Discipline of Botucatu Medical School, Univ Estadual Paulista – UNESP and also to the Post-Doctor Program of UNESP. This research received grants from Fundação de Amparo à Pesquisa de São Paulo (FAPESPNo 2008 ⁄ 09187-8; No 2009 ⁄ 01188-8) and Fundação de Amparo à Pesquisa de Mato Grosso (FAPEMAT No 735593 ⁄ 2008; No 453387 ⁄ 2009). The authors declare no conflict of interest and non-financial competing interests. References 1. American Dietetic Association. Position of the American Dietetic Association: promoting and supporting breastfeeding. J Am Diet Assoc 2009; 109: 1926–42. 2. Hanson LA. Session 1: feeding and infant development breastfeeding and immune function. Proc Nutr Soc 2007; 66: 384–96. 3. Paramasivam K, Michie C, Opara E, Jewell AP. Human breast Milk immunology: a review. Int J Fertil Womens Med 2006; 51: 208–17. 4. Lawrence RA, Lawrence RM. Breastfeeding: a guide for the medical profession. St Louis, MO: Mosby-Year Book, 2005: 105–70. 5. Newburg DS. Innate immunity and human milk. J Nutr 2005; 135: 1308–12. 6. Honorio-França AC, Carvalho MP, Isaac L, Trabulsi LR, Carneiro-Sampaio MMS. Colostral mononuclear phagocytes are able to kill Enteropathogenic Escherichia coli opsonized with colostral IgA. Scand J Immunol 1997; 46: 59–66. 7. França-Botelho AC, Honorio-França AC, França EL, Gomes MA, Costa-Cruz JM. Phagocytosis of Giardia lamblia trophozoites by human colostral leukocytes. Acta Paediatr 2006; 95: 438–43. 8. Islam N, Ahmed L, Khan NI, Huque S, Begun A, Yunus ABM. Immune components (IgA, IgM, IgG immune cells) of colostrum of Bangladeshi mothers. Pediatr Int 2006; 48: 543–8. 9. Abdulla EM, Zaidi FE, Zaidi A. Immune factors in breast milk: a study and review. Pak J Med Sci 2005; 21: 178–86. 10. Janoff EN, Fasching C, Orenstein JM, Rubins JB, Opstad NL, Dalmasso AP. Killing of Streptococcus pneumoniae by capsular polysaccharide-specific polymeric IgA, complement, and phagocytes. J Clin Invest 1999; 104: 1139–47. 11. Horta BL, Bahl R, Martines JC, Victora CG. Evidence on the long-term effects of breastfeeding. Systematic reviews and meta-analyses. World Health Organization Library Cataloguing-in-Publication, Geneva, 2007. 12. Taylor JS, Kacmar JE, Nothnagle M. A systematic review of the literature associating breastfeeding with type 2 diabetes and gestational diabetes. J Am Coll Nutr 2005; 24: 320–6. 13. Rudge MV, Calderon IM, Ramos MD, Maestá I, Souza LMS, Peraçoli JC. Perinatal outcome of pregnancies complicated by diabetes and by maternal daily hyperglycemia not related to diabetes. A retrospective 10-year analysis. Gynecol Obstet Invest 2000; 50: 108–12. 14. Mancini G, Carbonara AO, Heremans JF. Immunochemical quantitation of antigens by simple radial immunodiffusion. Immunochemistry 1965; 2: 235–6. 15. Novelli EL, Rodrı́guez NL, França EL, Gebra LMN, Ribas BO. High dietary carbohydrate and pancreatic lesion. Braz J Med Biol Res 1993; 26: 31–6. 16. Butte NF, Garza C, Smith EO. Variability of macronutrient concentrations in human milk. Eur J Clin Nutr 1988; 42: 345– 9. 17. Bitman J, Hamosh M, Hamosh P, Lutes V, Neville MC, Seacat J, et al. Milk composition and volume during the onset of lactation in a diabetic mother. Am J Clin Nutr 1989; 50: 1364–9. 18. Vanbeusekom CM, Zeegers TA, Martini IA, Velvis HJR, Visser GHA, Vandoormaal JJ, et al. Milk of patients with tightly controlled insulin-dependent diabetes-mellitus has normal macronutrient and fatty-acid composition. Am J Clin Nutr 1993; 57: 938–43. 19. Oliveira AMM, Cunha CC, Penha-Silva N, Abdllah VOS, Jorge PT. Interference of the blood glucose control in the transition between phases I and II of lactogenesis in patients with type 1 diabetes mellitus. Arq Bras Endocrinol Metabol 2008; 52: 473– 81. ª2010 The Author(s)/Acta Pædiatrica ª2010 Foundation Acta Pædiatrica 2011 100, pp. 550–556 555 Diabetes and changes human colostrum Morceli et al. 20. França EL, Nicomedes TR, Calderon IMP, Honorio-França AC. Time-dependent alterations of soluble and cellular components in human milk. Biol Rhythm Res 2010; 41: 333–47. 21. Wines BD, Hogarth PM. IgA receptors in health and disease. Tissue Antigens 2006; 68: 103–14. 22. Targowski SP. Role of immune factors in protection of mammary gland. J Dairy Sci 1983; 66: 1781–9. 23. L’Abbe MR, Friel JK. Superoxide dismutase and glutathione peroxidase content of human milk from mothers of premature and full-term infants during the first 3 months of lactation. J Pediatr Gastroenterol Nutr 2000; 31: 270–4. 24. Shimizu I, Makino H, Imagawa A, Iwahashi H, Uchigata Y, Kanatsuka A, et al. Clinical and immunogenetic characteristics of fulminant type 1 diabetes associated with pregnancy. J Clin Endocrinol Metab 2006; 91: 471–6. 25. Qin YD, Luo X, Huang DL, Xu CZ. Dynamic changes of enzymes activities and growth factors contents in human colostrums. Zhonghua Fu Chan Ke Za Zhi 2004; 39: 449–52. 556 26. Kasapovic J, Pejic S, Mladenovic M. Superoxide dismutase activity in colostrum, transitional and mature human milk. Turk J Pediatr 2005; 47: 343–7. 27. Saxena AK, Srivastava P, Kale RK, Baquer NZ. Impaired antioxidants status in diabetic rat liver effect of vanadate. Biochem Pharmacol 1993; 45: 539–42. 28. França EL, Feliciano ND, Silva KA, Ferrari CKB, HonorioFrança AC. Modulatory role of melatonin on superoxide release by spleen macrophages isolated from alloxan-induced diabetic rats. Bratisl Lek Listy 2009; 110: 517–22. 29. Maritim AC, Sanders RA, Watkins B. Diabetes, oxidative stress, and antioxidants: a review. J Biochem Mol Toxicol 2003; 17: 24–38. 30. Zarban A, Taheri F, Chahkandi T, Sharifzadeh G. Pattern of total antioxidant capacity in human milk during the course of lactation. Iranian J Pediatr 2007; 17: 34–40. ª2010 The Author(s)/Acta Pædiatrica ª2010 Foundation Acta Pædiatrica 2011 100, pp. 550–556