Solid-Phase Acid Extraction Improves Thiobarbituric Acid Method to Determine Lipid Oxidation S. RAHARJO, J.N. SOFOS, and G.R. SCHMIDT ABSTRACT Samples (11Og)of raw (17.2-22.6% fat) and cooked 12.6-16.4% fat) ground beef in plastic cups were stored aerobically at 4* 1°C. Lipid oxidation was measured by four versions of the thiobarbituric acid (TBA) test, including aqueous acid extractior& (TBA-C1s), direct heating, distillation, and unmodified aqueous acid extraction; and by sensory evaluation of rancid odor after 0, 2,4, 6, and 8 days storage. The TBA-Cl* method was more specific (P<O.O5) and its limit of dctcrmination was 20 times lower than the other methods in detecting malonaldehydc. Results correlated (r=0.856 to 0.883 in raw, and r = 0.936 to 0.981 in cooked meat) with sensory evaluation scores. Key Words: beef, TBARS, lipid, peroxidation, malonaldehyde INTRODUCTION SPECTROPHOTOMETRIC detection of the malonaldehyde- thiobarbituric acid (TBA) complex has been widely used for measuring lipid oxidation in meat products. The TBA test can be conducted in meat by: (a) directly heating the samples in the presence of TBA, followed by separation of the red pig- ment by centrifugation (Uchiyama and Mihara, 1978; Pokorny and Dieffenbacher, 1989); (b) distillation of the sample, fol- lowed by reaction of the distillate with TBA (Tarladgis et al., 1960; Rhee, 1978; Ke et al., 1984; Hoyland and Taylor, 1989); (c) extraction of the lipid portion of the sample with chloro- form-methanol and reaction of the extract with TBA (Witte et al., 1970; Salih et al., 1987). The direct heating and distillation TBA methods may form additional malonaldehyde and other TBA-reactive substances (TBARS) through degradation of pol- yunsaturated fatty acid hydroperoxides during analysis (Hoy- land and Taylor, 1991). A major concern with the TBA method on extracted material is that much of the preformed malonal- dehyde, naturally present in the aqueousphase of a meat sam- ple, may remain undetected (Schmedes and Holmer, 1989). Heating the lipid portion (containing different levels of poly- unsaturatedfatty acids) with the TBA solution could also gen- erate variable levels of TBARS including malonaldehyde (Gutteridge and Quinlan, 1983). Thus, that method may also provide information on susceptibility of different lipids to au- toxidation. The aqueous acid extraction TBA method uses milder con- ditions than the other tests. No heating is applied to the sample which minimizes formation of additional malonaldehyde and other TBARS. The procedure, however, is still not specific for detection of malonaldehyde, becauseother TBARS in the extract could have the same absorbance as the malonaldehyde- TBA complex (Squires, 1990; Draper and Hadley, 1990). We developed an aqueous acid extraction TBA-Cls method (Ra- harjo et al., 1992) with the potential of overcoming this inter- ference problem. However, the analytical reliability of that method needs to be compared with other TBA methods. Our objectives were to evaluate. the specificity and limit of determination of the TBA-C1s method compared to other tests. Authors Rahatjb, Sofos, and Schmidt are affiliated with the Dept. of Animal Sciences and Dept. of Food Science & Human Nutri- tion, Colorado State Univ., Fort Collins, CO 80523. Address in- quiries to Dr. J. N. Sofos. They involved direct heating, distillation and aqueousacid ex- traction. We also determined the correlation between results of the TBA-CIs analysis and other TBA methods. Finally, we evaluated the correlation behveen sensory evaluation of rancid odor and TBA numbers from the four TBA methods. MATERIALS & METHODS Sample preparation Frozen ground beef from a local grocery store, was thawed at 4°C overnight.-A portion (700g) of the ground beef in a beaker was cooked in water bath (National Aooliance Co.. Portland. ORI of 94 +-1°C for 20 min to in&al temper&re 7O”C, measurediy thermocouple (At- kins Technical Inc., Gainesville, FL). Samples (11Og) of raw (17.2- 22.6% fat) and cooked ground beef (12.6-16.4% fat) were placed in plastic cups,covered with caps (Solo Cup Co., Urbana, IL), and stored aerobically at 4kl”C. The extent of lipid oxidation was measured after 0,2,4,6, and 8 days storageby 4 different TBA tests and rancid odor development was assayedby sensory evaluation. TBA methods Direct heating. The direct heading TBA method was carried out as described by Uchiyama and Mihara (1978). Prior to homogeniza- tion, 0.15% of butylated hydroxytoluene (BHT) (Sigma Chemicals Co., St. Louis, MO) based on fat content, was added to each sample to prevent autoxidation during analysis (Pikul et al., 1983). The same amount of BHT was also added before homogenization to all samples analyzed by all other TBA procedures. Distillation. The distillation TBA procedure was performed as de- scribed by Tarladgis et al. (1960). Aqueous acid extraction. The aqueous acid extraction TBA test was performed as describedby Salih et al. (1987), except that 5% (w/ v) aqueous trichloroacetic acid (TCA) (Mallinckrodt, Paris, KY) was used for the extraction solvent. Aqueous acid extraction TBA-CIs. Ground beef samples (10 g) were homogenized with 40 mL of 5% (w/v) aqueous TCA in an Osterizer blender (Sunbeam Corp., Milwaukee, WI) at room temper- ature (~23°C) for 1 min. The meat homogenate was centrifuged at 10,000 x g for 5 min and the supernatant was filtered through a Whatman micro fiber glass filter grade C (Whatman, Hillsboro, OR) into a 50 mL volumetric flask. Filtrate volume was adiusted to 50 mL using 5% (w/v) aqueous TCA. A portion (5 mL) was reacted with 5 mL of 80 mM TBA in a test tube with a screw cap, while heating in a water bath of 94 f 1°C for 5 min. The pH of the formed red pigment was adjusted to 17 with 5N NaOH (Mallinckrodt, Paris, KY) and 0.2 mL of 3% (whr) ohosohate buffer of OH 7.2 (Becton Dickinson and Co., Cocke&viil& Mb) prior to pum’ping through a solid phase extraction Sep-Pak@ C18 cartridge (Waters, Milford, MA). Prior to its use, the C18 cartridge had been washed with 10 mL of absolute methanol (Mallinckrodt), followed by 10 mL of distilled water at ~20 mL/min. The flow rate was measuredby loading the solution into a 12-mL syringe (Becton Dickinson and Co.), connected to a C,,, cartridge, then manually pumped through the cartridge with a plunger for a specified period of time. The sample (10 mL) was loaded to the syringe and passedthrough the treated CIs cartridge at =5 mL/min to allow adequatetime for the C18 matrix to bind the red colored malon- aldehyde and other TBARS complexes. The eluted solution from the cartridge was discarded. Unreacted TBA solution and other compo- nents were removed by eluting the loaded sample with 10 mL of distilled water at ~10 mL/min. The eluted solution from the cartridge was discarded. The malonaldehyde-TBA complex was recovered and separatedfrom other TBARS by eluting the cartridge with 10 mL of absolute methanol at =lO mL/min. The absorbanceof the methanol Volume 58, No. 4, 1993-JOURNAL OF FOOD SCIENCE-921