The TPN admixtures prepared with 2/1 and 4/1 of calcium/phosphorus ratios were free of crystal formation, showing physicochemical compatibility during seven days of storage. Thus, no evidence of incompatibility between calcium and organic phosphorus was observed. It is recommended for preterm an intravenous administration of 200 to 800 mg/kg/day of calcium gluconate and 47 to 70 mg/kg/day of phosphate, concentrations closer to that observed in the uterus. These quantities are difficult to be obtained using the inorganic source due to the low solubility of these ions in aqueous solution [5, 25]. In order to maintain the biochemical and hormonal homeostasis and a suitable bone mineralization in preterm infants, the calcium/phosphorus ratio must be greater than 1. The inversion of this ratio can cause an increase in the secretion of parathormone, which leads to loss of phosphate in urine and osteopenia .
Clinical research has been showing the advantages in bone mineralization using organic phosphorus [4, 18, 26–28]. However, in addition to the concern of offering a clinical efficient TPN admixture, the physicochemical the stability of these formulations need to be considered and has been discussed [16, 29, 30]. Thus, in this work formulations were prepared varying the amount of calcium with fixed amount of phosphate to test the stability of organic phosphorus in the presence of high concentrations of calcium in admixtures containing lipid, vitamins and trace elements. Despite the TPN form a lipid emulsion, which may be subject to a process of physical instability, several metabolic benefits are associated with the administration of TPN admixtures over 2 in 1, such as the prevention of hyperglycemia and the essential fatty acids lack, less manipulation in the infusion line and a lower osmolarity of the final emulsion.
The physical stability of the lipid emulsions can be improved by using surfactants that form a coalescence energy barrier that carries electric charges around the dispersed liquid, aiming at decreasing the surface tension and increasing the repulsion force between the dispersed liquids . In the case of TPN admixtures containing egg lecithin as anionic emulsifier, the lecithin produces a negative charge around the lipid globules through the ionization of the phosphate groups. Any positively charged ion, such as calcium, could cause irreversible instability in this system, neutralizing the negative phospholipids droplets and promoting coalescence .
Therefore, the pH of the final admixture is important as a factor that directly influences the ionization of the lipid globule phospholipids, and also interferes in the dissociation of ions in solution. Many factors determine the final pH of the TPN admixture , as composition, final concentration of the amino acids, type and final concentration of the phosphate, cysteine addition, peroxidation and final glucose concentration. The pH for pediatric TPN admixtures ranges from 5.0 to 6.0, and the admixtures contain high concentration of amino acids such as HCl cysteine, histidine and arginine, as well as a high concentration of glucose [16, 32]. The pH of the admixtures studied remained within this range with no significant alterations over time, showing that the high calcium/phosphorus ratio in the admixture and the temperature of storage had no adverse effect on pH.
Among some of the clinical disorders associated with the parenteral infusion of an unstable lipid emulsion the production of hepatic accumulation of fat associated with oxidative stress, liver injury, and a low-level systemic inflammatory response can be mentioned . In this work it the stability of TPN admixtures with a high concentration of divalent ions was studied. The maximum globule size observed, shown in Table 4, represents a vanishingly small percentage when compared to the total numbers of globules measured, which is the reason because a percentage of this range cannot be visualized in Figure 2. The values observed were well below the recommended limit; however, the increased globules could occur sporadically, that is why it would be important to connect filters in line with the infusion equipment in order to ensure uniformity in the size of the lipid globules infused in these preterm infants. The TPN 1.2-μm filters significantly reduce the total number and concentration of enlarged fat globules and effectively remove rigid crystalline particulates . The other size parameters quantified through OM showed a profile similar to the mean diameter, with no significant alteration during the study. OM is a recognized technique for the qualitative evaluation of TPN admixture stability and its profile, since many randomly selected fields can be considered in the analysis. However, it fails to predict the quantitative evaluation of the lipid stability (PFAT5) of the admixtures overall due the small amount of sample employed in the OM analyses. So, these results need to be considered within the limitations of the technique and require confirmation by light obscuration.
Visual inspection is a simple method to evaluate the LE physicochemical compatibility, but extremely important. Color alteration, presence of precipitate and instability phases were evaluated in this step and confirmed microscopically. Color alteration in TPN is well known due to the Maillard reaction and oxidation of some vitamins [6, 14]. It is undesirable because it may represent a loss of bioavailability, due loss of some nutrient . In the present study the color alteration that occurred was unrelated to the concentration of calcium, since the PC admixture (without calcium) showed the same change as P2. Color alteration was observed in admixtures stored at 25°C and 37°C, but no alteration was observed when stored at 4°C. Color alteration is related not only to temperature, but also to light, which is the reason for the great difference in color between bags kept at 25°C, in room temperature, and at 4°C, in the refrigerator. In the bag stored at 25°C with photo protection, no color alteration was observed during the days of study (data not shown). Although the infusion of parenteral admixture must be at room temperature, these results indicate that the ideal storage condition is at 4°C, with photo protection, regardless of calcium. The data suggest that the admixtures should not be kept for a long period at 25°C before use, and that equipment with photo protection should be used.
The maintenance of the sterility of the admixtures studied in this paper was fundamental for us to attribute the alterations to possible incompatibility or lipid emulsion instability events due to the high calcium concentration, and not to changes caused by microbiological contamination.
In neonatology, for central access, the osmolarity level must be kept at around 800 mOsm/L, not exceeding twice the regular serum osmolarity [30, 35]. High osmolarity arises from high ion concentrations as well as glucose, and the ionic components may also affect the lipid globule stabilization. In this study, the experimental osmolarities were kept within the recommended range for the central access administration of the admixtures studied; and as expected, no significant differences in osmolarities were observed at different temperatures.
The infusion of oxidized lipids and secondary peroxidation products can be extremely cytotoxic, and may cause many disorders, such as hepatic steatosis , hyper triglyceridemia , increase in vascular resistance in the lung [17, 38], lung remodeling  and chronic lung diseases of prematurity [40, 41]. TPN is a potential source of oxidants and this is particularly dangerous in preterm infants who are vulnerable to oxidative stress [42, 43]. The peroxide formation caused by exposing TPN admixtures to phototherapy is well-known in the neonatology field [23, 44, 45]. This is corroborated by the demonstration that physical protection of the light incidence in TPN admixtures leads to formation of even lower concentrations of peroxides than the usual photo protection [23, 45].
The objective of this study was to evaluate the stability of the formulations with high calcium/phosphorus ratio, so experiments were performed varying only the temperature of storage without light exposure. Despite the fact that all peroxide values measured were very low (Figure 3), the admixtures kept photo-protected and totally in absence of light, presented even lower peroxide values, different from those observed in Figure 3A. This observation reinforces the necessity of photo protection for the TPN admixture infusion equipment.
The calcium and phosphorus presence even at high concentrations did not show a statistically significant change in the peroxides value when compared with the control formulation. Low values of peroxide were measured, comparable to values previously demonstrated by other authors when the formulation was totally photo protected [23, 45]. The low peroxide value observed could be attributed to many favorable factors, such as the multilayered bag used to store the formulations [42, 46]. It has been shown that TPN admixtures in multilayered bag present less oxidation evaluated by hydroperoxide formation [46, 47]. This bag has low oxygen permeation and could be collapsed to eliminate, as much as possible, the oxygen in the bag. Moreover, the presence of amino acids [42, 48] in formulations containing trace elements, lipids and vitamins  decrease the peroxide formation and its clinical complications [39, 42]. Lavoie et al  showed that during the 6 hours after TPN preparation the balance between generation and consumption of peroxides is positive, and after this time they presume that the balance changed in favor of the consumption, during their transformation into free radicals via the Fenton-like reaction induced by trace elements present in the PN solution .
Although it was shown that the vitamin solutions cause more peroxidation [36, 41, 42], some vitamins have been shown to have antioxidant effects, as vitamin C [23, 39, 45, 50]. The oxidation of aminoacids and vitamin C could prevent the lipid peroxidation turning the peroxidation undetectable. When the amino acid oxidizes the formulation becomes darkened, as observed in admixtures stored at 25°C exposed to artificial light. The relative protective effect of amino acid in the peroxidation in the presence of vitamins and light for 24 hours was already demonstrated .
Antioxidant activity of trace elements is demonstrable in some constituents as selenium and copper [51, 52]. However, by evaluation of peroxide in the urine in preterm infants, it was observed that the use of trace elements did not affect peroxide production . Even though, Steger et al  had shown by in vitro studies the increase in peroxidation in TPN admixtures associated with trace elements, these results may be due to the long period of analysis up to 30 days .