Liquid chromatographic determination of vitamin D in infant formula

A method is described for the determination of vitamins D2 + D3 in milk- and soy-based infant formula. Vitamins D2 and D3 are extracted from the saponified sample and converted to isotachysterols with acidified butanol. Reverse-phase liquid chromatography (LC) is used to remove interferences, and total vitamin D is quantitated using normal-phase LC. Recoveries of spiked samples averaged 97.6% for milk-based infant formula, and 98.8% for soy-based infant formula. This method quantitates vitamin D2 + D3 in infant formulas containing as low as 40 IU/qt when prepared according to label direction.     

Liquid chromatographic determination of vitamins D2 and D3 in fortified milk and infant formulas

A liquid chromatographic (LC) method was developed for determining vitamins D2 and D3 in fortified milk and infant formulas. The lipid-soluble components were extracted from the aqueous phase by homogenizing in isopropanol-methylene chloride with magnesium sulfate added to remove water. The vitamins were fractionated from the lipid material by using gel permeation chromatography (GPC) followed by further cleanup of the combined GPC fractions on a muBondapak/NH2 column. Four muStyragel (100 A) columns connected in series were used for GPC fractionation of sample extracts in methylene chloride. Injection and collection were repeated 3 times to collect enough vitamin D for quantitation. The muBondapak/NH2 column, using a mobile phase of methylene chloride-isooctane-isopropanol (600 + 400 + 1), resolved vitamin D from other UV-absorbing compounds and soy sterols in infant formula and from cholesterol in milk. Vitamins D2 and D3 coeluted as one peak, with the resolution and vitamin level sufficient for visual monitoring (280 nm/0.02 absorbance unit full scale) in a collection time of 22-26 min. A Zorbax ODS (6 micron) column and a methylene chloride-acetonitrile-methanol (300 + 700 + 2) mobile phase were used for LC quantitation; vitamins D2 and D3 were baseline resolved in about 11 min. The infant formula samples included ready-to-use and concentrated liquids prepared in nonfat milk base or soy base fortified with vitamins D2 or D3 at 400 IU/qt or L (10 micrograms). The mean percent recovery of added vitamin D3 (400-500 IU/qt) from infant formula (n = 7) was 89.6 +/- 6.7 (coefficient of variation (CV) 7.5%).(ABSTRACT TRUNCATED AT 250 WORDS)     

Sample preparation and liquid chromatographic determination of vitamin D in food products

Vitamin D in different fortified foods is determined by using liquid chromatography (LC). Sample preparation is described for fortified skim milk, infant formulas, chocolate drink powder, and diet food. The procedure involves 2 main steps: saponification of the sample followed by extraction, and quantitation by LC analysis. Depending on the sample matrix, additional steps are necessary, i.e., enzymatic digestion for hydrolyzing the starch in the sample and cartridge purification before LC injection. An isocratic system consisting of 0.5% water in methanol (v/v) on two 5 microns ODS Hypersil, 12 X 0.4 cm id columns is used. Recovery of vitamin D added to unfortified skim milk is 98%. The results of vitamin D determination in homogenized skim milk, fortified milk powder, fortified milk powder with soybean, chocolate drink powder, and sports diet food are given.     

Liquid chromatographic determination of vitamin K1 trans- and cis-isomers in infant formula

The normal phase liquid chromatographic (LC) method for determination of trans- and cis-isomers of vitamin K1 (phylloquinone) in infant formula described here uses an Apex silica column, isocratic elution, and UV absorption detection at 254 nm. Vitamin K1 is extracted quantitatively from the product matrix by pretreating the as-fed liquid with concentrated ammonium hydroxide and methanol, and then extracting it with a 2:1 mixture of dichloromethane and isooctane. The extract is cleaned up by silica open-column chromatography and concentrated for LC analysis. For trans-vitamin K1, the method precision is less than or equal to 3.3% RSD (relative standard deviation), and the spike recovery is 98 +/- 4%. For cis-vitamin K1, the precision is less than or equal to 12% RSD, determined at levels near the detection limit, and the spike recovery is 95 +/- 9%. The detection limit is 0.3 ng for both isomers at signal/noise = 3.     

Determination of taurine in infant formulas using ultrafiltration and cation-exchange chromatography

A fast and simple method for determination of taurine in infant formulas has been developed. The sample preparation uses disposable ultrafiltration cartridges to remove protein and clarify the sample. Hydrolysis is avoided, simplifying the procedure and increasing efficiency. One mL sample is centrifuged in a cartridge for 45 min. The filtrate is diluted with pH 2.2 citrate buffer and injected into a high performance amino acid analyzer. A cation-exchange column (sodium phase) is used with a single buffer eluant and an isocratic chromatographic program. Colorimetric detection is performed following post-column ninhydrin reaction. Chromatographic resolution from other ninhydrin-positive compounds is excellent. Average recoveries for 3 levels of spike for various products were 100-102%. Precision is 1-3% RSD, depending on product. Linearity, specificity, and ruggedness are excellent. The method is applicable to quality control testing of milk-based, soy-based, and prehydrolyzed protein-based infant formulas in the ready-to-use, concentrate, and powder forms. A variety of commercially available infant formulas from different manufacturers were analyzed and all were found to contain taurine levels comparable to human milk. Some human milk and cow's milk samples were also analyzed and results compare well with literature values.     

Vitamin A and vitamin E content of infant formulas produced in the United States

Vitamin A (vitamin A palmitate) and vitamin E (alpha-tocopheryl acetate) levels were determined in 77 samples of fortified infant formulas manufactured by 4 firms in the United States from 1981 to 1983 and were compared by formulation base (soy, milk) and manufacturing firm. For vitamin A and vitamin E, the mean values (IU/100 kcal) were 454 +/- 95 (range 248-614) and 2.0 +/- 0.7 (range 1.1-5.0), respectively. No significant differences (alpha = 0.05) were found in levels (IU/100 kcal) of vitamin A and vitamin E between milk- and soy-based formulas. When the mean vitamin A and vitamin E levels of formulas produced by the various firms were compared on an IU/100 kcal or percent of label declaration basis, significant differences (alpha = 0.05) were found among firms. Mean vitamin A levels for the various products compared to label declarations ranged from 126% of declared for the ready-to-use formulas to 139% of declared for the powders. Mean vitamin E levels ranged from 97% of declared for ready-to-use formulas to 118% of declared for concentrates. Except for one sample that contained 248 IU vitamin A/100 kcal, the formulas met the requirements of the 1980 Infant Formula Act.     

Fast and simple liquid chromatographic determination of nonphosphorylated thiamine in infant formula, milk, and other foods

A very fast and simple method for determination of nonphosphorylated thiamine in infant formula products, milk, and other nonfortified foods using reverse-phase ion-pairing liquid chromatography (LC) has been developed. Sample preparation consists of merely acid treatment to precipitate protein, followed by gravity filtration. No concentration, extraction, derivatization, or preliminary column cleanup is necessary. The chromatography is done on muBondapack C18 with an aqueous mobile phase containing 0.15% sodium hexane sulfonate, 20% MeOH, 1.5% HOAc, and 0.1% EDTA at a flow rate of 2.5 mL/min. Ultraviolet detection at 248 nm is used. A typical run takes 7 min, and 60 samples can be processed in 4 h. Results average from 96 to 104% of theory for the infant formula products analyzed. A 99 to 103% recovery of spike has been demonstrated. Method precision is good (2 to 4% RSD, short-term, and 2 to 5% RSD, long-term, depending on sample type). Peak separation from thiamine phosphate esters is achieved. Specificity is demonstrated by UV spectral scan and absorbance ratios. Equivalency to a microbial method (validated against the official AOAC fluorometric method) was established. The method is used for high-volume quality control testing of milk-based infant formula products in the ready-to-use, concentrate, or powder form.     

Liquid chromatographic cleanup and determination of low levels of vitamin D3 in sheep plasma

A liquid chromatographic (LC) method is described for the determination of vitamin D3 in sheep plasma. Samples are extracted by one of 2 different methods, depending on the concentration of vitamin D3. The samples are purified by using either a Sep-Pak silica cartridge or a small alumina column, followed by additional cleanup on a Metalsorb LC column. Final analysis was carried out on a 5 micron C18 column using a radial compression separation system with an acetonitrile-methanol solvent system. Vitamin D3 was completely resolved from any interfering compounds in the plasma; total run time was less than 15 min, using a variable wavelength detector set at 264 nm. The method was successfully applied to samples at levels of 1-10 ng added vitamin D3 mL sheep plasma, with recoveries in the range 90-97%.     

Normal-phase liquid chromatographic determination of menadione in animal feeds

A simple method is presented for determination of menadione in vitamin premixes and feedstuffs by normal-phase liquid chromatography (LC). Vitamin K3 is extracted and converted to free menadione, which can be determined directly by LC analysis. Peak area or height is measured at 251 nm, and menadione is quantitated by comparison with the working standard. Menadione can be estimated with a detection limit of 2.5 ppm. Recoveries for premixes ranged from 97.3 to 98.3% and for feedstuffs from 93.7 to 96.8%. The method allows the assay of all commercial K3 compounds in pure or stabilized form and is applicable to a wide variety of feeds and premixes.     

Amino acid rating method for evaluating protein adequacy of infant formulas

Amino acid profiles and/or protein digestibility (by the rat balance method) were determined for various forms (powder, ready-to-use, liquid concentrate, etc.) of cow's milk- and soy-based infant formulas obtained from 4 manufacturers. The essential amino acid data of the formulas were compared with that of human milk for the calculation of amino acid scores (based on the single most limiting amino acid). The product of amino acid score and total protein (g/100 kcal) was then termed "amino acid rating." Amino acid scores for the milk- and soy-based formulas ranged from 59 to 90 and from 59 to 81%, respectively, due to deficiencies in sulfur amino acids and/or tryptophan. Because of significantly higher total protein contents (g/100 kcal) of soy- (2.65-3.68) and milk-based (2.20-2.95) infant formulas compared to human milk (1.5), the relative amino acid ratings (human milk = 100) for all infant formulas except 2 liquid concentrates (having values of 87%) were above 100%. Values for true digestibility of protein in milk- and soy-based formulas ranged from 87 to 97 and from 92 to 95%, respectively. When corrected for protein digestibility, the relative amino acid ratings for all the milk-based liquid concentrates were below 100% (77-98%).     

Liquid chromatographic determination of cholecalciferol in rodent baits

Cholecalciferol (vitamin D3) is extracted in acetonitrile on a Goldfisch apparatus, diluted to volume, and determined by reverse phase liquid chromatography (LC). The sum of the peak heights of pre-vitamin D3 and cis-vitamin D3 is used for quantitation. The method was tested for precision, linearity, and recovery. Quadruplicate analyses of 5 formulation samples gave relative standard deviations of 1.56-2.65%. Linearity was excellent with regression and correlation coefficients of 0.9997 and 0.9998, respectively. Recovery was 98.0 +/- 2.7%. The method is applicable to 0.075% cholecalciferol rodent baits.     

Screening method for vitamins A and D in fortified skim milk, chocolate milk, and vitamin D liquid concentrates

Vitamin A was determined in fortified chocolate milk and skim milk; vitamin D was determined in fortified chocolate milk, skim milk, and vitamin D concentrates, using reverse phase high pressure liquid chromatography (HPLC). The sample is saponified, extracted with hexane, and chromatographed in an HPLC system on a 10 micron Vydac TP reverse phase C18 column, using acetonitrile-methanol (9+1) as the mobile phase. For 6 replicates, the recoveries of vitamins A and D, using this procedure, were 99 and 98%, respectively.     

Vitamin D3 fortification, quantification, and long-term stability in Cheddar and low-fat cheeses

Considering the widespread insufficiency of vitamin D, the fortification of additional foods with vitamin D is warranted. The objective of this research was to assess the feasibility of vitamin D3 fortification in natural hard cheeses. We examined the recovery, distribution, long-term retention, and heat stability of the vitamin in industrially made fortified Cheddar and low-fat cheeses. The results indicated that the vitamin D3 did not degrade during processing, over 1 year of ripening (3-8 degrees C), or after thermal treatment at 232 degrees C for 5 min. Vitamin D3 recovery in the fortified Cheddar and low-fat cheeses were, respectively, 91 and 55% of the vitamin D3 added to the milk used to make each cheese. The remaining vitamin D3 was entrained in the whey. The vitamin D3 was uniformly distributed throughout the blocks of cheese. The fortification process did not alter the yield, chemical composition, or flavor of the Cheddar cheese. We conclude that industrially manufactured Cheddar and low-fat cheeses are suitable for vitamin D3 fortification.     

Analysis of fat-soluble vitamins. XXIII. High performance liquid chromatographic assay for vitamin D in vitamin D3 and multivitamin preparations.

Vitamin D is determined in preparations containing other fat-soluble vitamins by high performance liquid chromatography (HPLC). The unsaponifiable residue is extracted and separated from interferences by reverse phase chromatography; the fraction corresponding to vitamin D3 is collected and quantitated using normal phase chromatography (amylalcohol-n-hexane as mobile phase) by measuring the vitamin D3 and pre-vitamin D3 peaks at 254 nm. Previtamin D3 content is calculated as vitamin D3 with a conversion factor (determined on the equipment used). Application of the method to vitamin AD3 mixtures in oils gives 98-102% recovery. The reproducibility, using an external standard, is 2-3%, calculated as the coefficient of variation; with an internal standard, the coefficient of variation is 1-1.5%. The method measures potential vitamin D3 content in preparations containing greater than or equal to 200 IU/g in the presence of all known vitamin D3 isomers, vitamin A, and vitamin E.     

Liquid chromatographic determination of vitamin D in fortified milk

A method for the determination of vitamins D2 + D3 in fortified milk is described. Vitamins D2 and D3 are extracted from the saponified sample and converted to isotachysterols with antimony trichloride. The isotachysterols are quantitated using liquid chromatography with ultraviolet detection at 301 nm, which is the absorption maximum. At this wavelength other materials present in the sample do not interfere with the analysis of isotachysterols and therefore a cleanup step is avoided. Recoveries of vitamin D added to skim milk were 98.1% (SD 5.3), 96.7% (SD 3.3), and 96.0% (SD 5.1) for samples fortified with 200, 400, and 600 IU/quart, respectively. For whole milk, recoveries were 102.0% (SD 6.5) and 97.1% (SD 3.5) in samples fortified with vitamin D equivalent to 200 and 400 IU/quart, respectively. The detection limit for vitamin D is 40 IU/quart.     

Reverse phase high pressure liquid chromatographic determination of vitamin K1 in infant formulas

A reverse phase high pressure liquid chromatographic (HPLC) method for quantitating vitamin K1 in enzymatic hydrolysates of infant formula is described. The vitamin is extracted with n-pentane before determination by isocratic and isothermal reverse phase HPLC. Recovery of vitamin K1 added to 5 infant formulas ranged from 84 to 103%.     

Analysis of fat-soluble vitamins. XXIX. Liquid chromatographic determination of vitamin D in AD concentrates: collaborative study

A simplified liquid chromatographic (LC) method for determining vitamin D in vitamin AD concentrates (greater than or equal to 5000 IU vitamin D/g) was collaboratively studied. In the simplified method, the 2 columns specified in AOAC LC method 43.101-43.109 are replaced by a single column, which separates the vitamin D isomers and the vitamin A esters. The procedure for oils includes dissolution and quantitation by normal phase LC. Dry multivitamin concentrates and aqueous dispersions are treated with an enzyme system and the vitamins are extracted with n-pentane. Six coded samples were distributed to 16 laboratories; 15 collaborators returned their results. Estimates of repeatability and reproducibility for the oil samples were 1.1 and 3.1%, respectively; for the high-level concentrated dry preparation 1.4 and 3.9% and for the low-level concentrated dry preparation 1.3 and 11.4%, respectively. These values are a considerable improvement over the results obtained in the 1979 multivitamin collaborative study. The method has been adopted official first action for determination of vitamin D in vitamin AD concentrates containing greater than or equal to 5000 IU vitamin D/g.     

Liquid chromatographic analysis of riboflavin vitamers in foods using fluorescence detection

The analysis of free riboflavin (RF) and its two coenzymes, flavin mononucleotide (FMN) and flavin-adenine dinucleotide (FAD), is optimized using reversed phase liquid chromatography with fluorescence detection. The stationary phase was amide-based and endcapped with trimethylsilyl, and the isocratic mobile phase consisted of a 10:90 v/v acetonitrile/phosphate buffer (pH 5). Peaks were identified by the retention characteristics and fluorescence spectra. Detection limits were 0.03, 0.05, and 0.24 ng for RF, FMN, and FAD, respectively. The vitamins were extracted using acetonitrile and the phosphate buffer. The procedure was applied to the determination of B2 vitamers in different types of food such as milk and soy-based infant formulas, beer, fruit juices, and honey of different types. Most B2 vitamin appeared as RF, while the coenzymes were present in lower amounts. The method was validated using two certified reference materials, and results within the certified range were obtained.     

Calcium, iron, and zinc uptake from digests of infant formulas by Caco-2 cells.

Our aim was to estimate the bioavailability of calcium, iron, and zinc from infant formulas using a model that includes in vitro digestion and a Caco-2 cell culture to estimate the uptake. The cell culture conditions were selected, and uptake assays were carried out first with calcium, iron, and zinc standard solutions, and then with the soluble fraction of enzymatic digests of an adapted milk-based and a soy-based infant formula. It was not possible to measure the uptake of calcium, iron, and zinc from standard solutions added to the cell cultures in amounts similar to those present in infant formula digests with our method. The fact that it was, however, possible in the case of enzymatic digests suggests the presence of components in the digests that enhance mineral uptake. When mineral uptakes were expressed as percentages of the mineral present, statistically significant differences were found in the case of calcium between the uptake from the milk- and the soy-based formulas. For iron and zinc no such differences were observed.     

Evaluation of infant formula protein quality: comparison of in vitro with in vivo methods

Two-week protein efficiency ratio (2-wk PER), net protein ratio (NPR), calculated-protein efficiency ratio (C-PER), and discriminant computed-protein efficiency ratio (DC-PER) of milk- and soy-based infant formulas were compared to a 4-week protein efficiency ratio (PER). Expressed relative to ANRC casein, 2-week PER and NPR correlated significantly (P less than 0.01, r = 0.90) with PER. Although C-PER and DC-PER also correlated significantly (P less than 0.01) with PER, r = 0.71 and r = 0.87, respectively, these in vitro methods did not distinguish differences in protein quality among soy-based infant formulas. C-PER and DC-PER, as currently designed, are not applicable to the measurement of protein quality for all types of infant formulas.