Freeze-Thaw and Cooking Effects on Broiler Breast Fillets with Extreme Initial L* Values 1

Freeze-Thaw and Cooking Effects on Broiler Breast Fillets with Extreme Initial L* Values 1 J. Galobart 2 and E. T. Moran, Jr. 3 Poultry Science Department. Auburn University, Auburn, Alabama 36849 ABSTRACT Five hundred broiler males were grown to cooked (internal temperature 80 C). Thawing reduced the 56 d and processed under common terms. Front halves were deboned 24 h postmortem to obtain breast fillets, and CIELAB light reflectance was measured on the skin side of each fillet 24 h later. All fillets were bagged and frozen ( 20 C) for 5 mo. Then the fillets exhibiting the lowest (dark), median (normal), and highest (pale) L* values 48 h postmortem were thawed (3 d at 4 C) and L* value in the pale fillets and increased it in the dark ones, and cooking further increased L* value and reduced the differences in L*, a*, and b* between groups. Thawing and cooking losses were not affected by initial L* value until they were combined. Total losses increased with initial L*, which was in parallel with a lower increase in thickness after cooking. (Key words: breast fillet, fillet quality, light reflectance, thawing loss, cooking loss) 2004 Poultry Science 83:2093 2097 INTRODUCTION One of the most important problems in the poultry processing industry is the development of the pale, soft, and exudative (PSE)-like condition. Pale, soft, and exudative-like meat is caused by a rapid decrease in ph early postmortem when carcass temperatures are still elevated. The combination of decreased ph and elevated temperature results in protein denaturation, which can lead to low water-holding capacity, pale color, and soft texture. It can cause problems during cooking by increasing water losses, resulting in increased yield loses, poor meat binding and, dry, soft texture. Evidence shows that a strong negative correlation exists between poultry breast muscle lightness (L*), ph, and water-holding capacity (Barbut, 1993, 1996; McCurdy et al., 1996; Allen et al., 1997, 1998; Liu and Van Laack, 1998; Fletcher, 1999; Le Bihan-Duval et al., 1999). In fact, drip and cooking losses have been reported to be higher when L* increases (Allen et al., 1998; Owens et al., 2000; Woelfel et al., 2002). Woelfel et al. (2002) suggested that L* value may be a better predictive tool than ph for sorting fillets for potential functionality. 2004 Poultry Science Association, Inc. Received for publication February 24, 2004. Accepted for publication September 3, 2004. 1 Data in part were presented at the 2003 XVI European Symposium on the Quality of Poultry Meat, Ploufragan-St. Brieuc, France. 2 This author was supported by a Fulbright-Generalitat de Catalunya grant. 3 To whom correspondence should be addressed: emoran@ acesag.auburn.edu 4 HunterLab, Reston, VA. A high proportion of breast meat is commercialized in frozen form, and it has to go through thawing and cooking before consumption. Although the relationship between L* and water losses during cooking has been established, the alterations to breast meat during freezing and thawing have received little attention. Thus, the objective of the present work was to recover breast fillets from an entire population of broiler males having a controlled background, and use those fillets having median and extreme L* values for evaluation of changes with freeze-thawing and cooking. MATERIALS AND METHODS Five hundred broiler males (Ross Ross 308) were raised to 8 wk of age under common feed and management. At 56 d of age, the broilers were processed and carcasses were maintained at 4 C for 24 h. At that point, breast fillets (pectoralis major) were removed for experimentation. Light reflectance (CIE L*a*b* scale) was measured 24 h later (48 h postmortem) on fillets originating from the right side of each carcass using a Miniscan XE 4 using illuminant D65 and standard 10 observer. All readings were taken at the center of the muscle surface that had been adjacent to the skin. The fillets were then placed in plastic bags and frozen ( 20 C). The fillets were sorted depending on L* value, and after 5 mo of frozen storage, those having the highest (pale), lowest (dark), and median (normal) L* values relative to the total population (33 fillets per category) were selected for experimentation. Abbreviation Key: PSE = pale, soft, and exudative. 2093

2094 GALOBART AND MORAN TABLE 1. Description of breast fillets selected from the population of broiler males at 56 d of age based on L* value for measurement of freeze-thaw and cooking characteristics L* a* b* L* category 1 Mean Range Mean Range Mean Range Dark 56.8 51.7 to 58.3 7.5 4.7 to 14.1 14.3 8.5 to 19.6 Normal 62.9 62.8 to 63.1 5.9 3.8 to 8.8 15.3 12.1 to 19.8 Pale 68.4 66.9 to 73.2 5.5 2.8 to 11.8 16.3 8.5 to 25.2 1 Each category was represented by 33 fillets selected from a total population of 497. Those fillets were individually weighed, wrapped in paper towels, placed inside plastic bags, and allowed to thaw for 3 d at 4 C in a walk-in refrigerator. Each thawed fillet was subsequently weighed, its maximum length, width, and thickness determined, and light reflectance measured at the same location. Thawed fillets were cooked in a convection oven preheated at 163 C, until the internal temperature was 80 C, as measured by indwelling thermocouples, and cooking time was recorded. Cooked fillets were allowed to cool for 45 min before weight, dimensions, and color were measured again. For the cooked meat color determination, approximately onethird of the thickness of the fillet was removed from the skin side to expose an area of meat free from cookingrelated discoloration. Fillets were then freeze-dried to obtain the residual free water content. The same person conducted respective measurements at each stage of experimentation. Thawing and total losses were calculated as the percentage of weight difference of the thaw and cooked fillets from the raw weight, respectively. Cooking losses were calculated as the percentage of weight difference between the cooked weight and the thaw weight. Data regarding weight and light reflectance measurements on raw, thawed, and cooked samples, and dimensions in thawed and cooked samples were analyzed as repeated measures using the MIXED procedure of SAS (SAS Institute, 2001). Thawing, cooking, and total losses as well as change in dimensions were evaluated by AN- OVA in a completely randomized design. Mean separation procedure was performed by orthogonal polynomial techniques. Pearson s correlation coefficients and probabilities were generated using the Correlation procedures of SAS (SAS Institute, 2001). RESULTS AND DISCUSSION Description of the Population of Fillets Used The fillets originating from a population of 500 male broilers had a mean weight and standard error of 672 ± 3.1g, which represents 24.7 ± 0.07% of carcass weight. Mean light reflectance values were as follows: L* = 62.9 ± 0.52; a* (redness) = 22 ± 0.27 and b* (yellowness) = 15.6 ± 0.44. A normal distribution of L* values existed with the entire population of fillets that were common in strain source, sex, terms of production, and conditions of processing. The values found in this study were sensibly higher than those published by other authors using the same CIE L*a*b* scale. Those differences may be due to the time of reading (hours postmortem), differences in the illuminant and observer used in the spectrocolori- TABLE 2. Variation in light reflectance among fillets having selected values from raw through thawed and cooked state 1 L* category Raw-thaw Thaw-cooked Overall L* Dark 1.46 a 25.23 a 26.70 a Normal 0.09 b 22.42 b 22.33 b Pale 2.42 c 19.54 c 17.11 c SEM 0.467 0.489 0.308 P *** *** *** a* Dark 2.42 a 7.17 a 4.75 a Normal 2.03 ab 5.78 b 3.75 b Pale 1.54 b 4.91 c 3.36 b SEM 0.217 0.237 0.233 P * *** *** b* Dark 3.71 0.42 b 3.29 a Normal 3.32 2.04 a 1.28 b Pale 3.43 2.68 a 0.75 b SEM 0.308 0.287 P NS *** *** a c Values within columns for each factor without common superscript are significantly different (P 0.05). 1 Each value is represented by measurements on 33 fillets from each L* category when raw. *P 0.05, ***P 0.001.

FREEZE-THAW, COOKING, AND FILLET PALENESS 2095 TABLE 3. Pearson correlation coefficients (r) and probabilities of raw, thawed, and cooked broiler breast meat lightness (L*), redness (a*), and yellowness (b*) Raw Thawed Cooked L* a* b* L* a* b* L* a* b* Raw L* 1 a* 0.468*** 1 b* 0.303** 0.382*** 1 Thawed L* 0.778*** 0.360*** 0.452*** 1 a* 0.605*** 0.785*** 0.069 NS 0.621*** 1 b* 0.319** 0.232* 0.755*** 0.340*** 0.128 NS 1 Cooked L* 0.496*** 0.528*** 0.183 NS 0.442*** 0.615*** 0.102 NS 1 a* 0.333*** 0.696*** 0.330*** 0.282** 0.630*** 0.363*** 0.571*** 1 b* 0.140 NS 0.641*** 0.560*** 0.102 NS 0.514*** 0.587*** 0.379*** 0.834*** 1 *P 0.05, **P 0.01, ***P 0.001. meter, or age of the birds (Abeni and Bergoglio, 2001; Petracci and Fletcher, 2002). Moreover, most reported studies perform light reflectance readings on the boneside of the breast fillet, whereas in our study it was performed on the skin-side surface of the muscle. Unpublished results from previous experimentation has shown that L* is higher by an average of 5 units on the skin-side than on the bone-side of the breast fillet. Despite the fact that the absolute mean values of the population in our study are 7.7 L* units higher than those reported by Wilkins et al. (2000) (62.9 vs. 55.2), the range found between the palest and the darkest fillet in both studies is comparable (21.5 vs. 22.3). After sorting all fillets according to L* value, the fillets (33 in each category) having the median and extreme low and high values led to fillets with maximal differences in L* values uncomplicated by background (Table 1). These fillets were used to examine the significance of L* to changes associated with freezing, thawing, and cooking. L* values were clearly different between the 3 categories, with a difference of 11.6 L* units between the mean of the pale and dark groups, which gave us enough difference to study the relationship of meat color with the changes that occur during freeze-thawing. Unavoidably, the variation in L*, a*, and b* was more extensive in the dark and pale categories than in the median category. Changes During Freeze-Thawing and Cooking Light Reflectance. Light reflectance is an easy procedure to perform and L* value is commonly used to detect fillets having a PSE-like problem. Thawing caused a reduction in L* values of pale and an increase in dark fillets, but did not modify L* values of normal fillets (Table 2). An increase in a* and b* was observed after thawing. Whereas the increase in b* was independent of the L* category, the increase in a* was lower for the pale than for the dark fillets. Cooking led to a further increase in L* with all fillets from the thawed state that was, again, higher as the original L* decreases. Thus, the overall effect of freeze-thawing and cooking was an increase in L* values that was inversely proportional to the initial L* value of the breast meat 48 h postmortem. A reduction in a* values was observed after cooking, being more acute in the dark fillets than in the pale and normal ones. On the contrary, reduction in b* during cooking was higher as the initial L* value increases. Thus, while the net effect of freeze-thawing and cooking on a* and b* are opposed (a* is reduced and b* increased), the magnitude of both changes is higher in the dark fillets than in the pale ones. However, the effect of thawing on light reflectance parameters was much less than the effect of cooking. Little TABLE 4. Fillet weight at different states and its loss during freeze-thawing and cooking 1 Fillet weight (g) Weight loss (%) L* category Raw Thawed Cooked Thaw 2 Cook 3 Total 4 Dark 319.8 288.8 216.2 9.7 25.1 32.4 b Normal 324.5 291.5 215.2 10.2 26.2 33.7 ab Pale 331.7 296.0 215.7 10.8 26.9 34.9 a SEM 7.41 7.41 3.35 0.23 0.38 0.34 P NS NS NS NS NS * a,b Values within a column without common superscript are significantly different (P 0.05). 1 Each value is represented by measurements on 33 fillets from each L*-category. 2 Thawing loss from raw state. 3 Cooking loss from thaw state. 4 Thawing plus cooking losses from raw state. *P 0.05.

2096 GALOBART AND MORAN TABLE 5. Change in physical dimensions from thawed to cooked state of fillets having different L* values measured when raw 1 Thawed dimensions (mm) Cooked dimensions (mm) Variation (%) L* category Length Width Thickness Length Width Thickness Length Width Thickness Dark 205.7 112.3 24.1 154.8 96.3 30.5 24.6 14.0 27.6 a Normal 205.9 112.8 23.2 156.0 95.8 29.6 24.1 15.0 28.2 a Pale 202.3 112.4 24.5 151.8 93.7 29.3 24.9 16.5 20.5 b SEM 1.11 0.73 0.32 0.87 0.69 0.34 0.48 0.58 1.35 P NS NS NS NS NS NS NS NS * a,b Values within a column without common superscript are significantly different (P 0.05). 1 Each value is represented by measurements on 33 fillets from each L* category. *P 0.05. information exists on fillet L* values beyond the fresh state. Lyon et al. (1976) did not observe differences in Hunter L values of fillets between samples kept fresh and those frozen 6 d then thawed at 2 C overnight. In the present study, subsamples at 48 h postmortem having L* values in the median group did not change from fresh values with freezing and thawing. Overall, research generally supports that L* values of fillets increase through the first 24 h postmortem with little change thereafter; however, samples involved random representatives from a population (McCurdy et al., 1996; Qiao et al., 2002). Cooking is known to greatly increase fillet L* values (Yang and Chen, 1993; Lyon et al., 1998, 2003; Fletcher et al., 2000; Claus et al., 2001). Le Bihan-Duval et al. (1999) also described an increase in a* and b* values in breast meat stored for 3 d at 4 C. The net effect of freezing, thawing, and cooking was to reduce difference in L*, a*, and b* values between extremes within the whole population, which agrees with the results of Fletcher et al. (2000). The discoloration observed after thawing and cooking may be related to the breakdown of metmyoglobin and sulfomyoglobin into small molecules, probably due to the action of specific enzymes, which are released from the frozen meats and activated during thawing (Liu and Chen, 2001). Raw meat color values resulted in highly significant positive correlations with all of the corresponding thawed and cooked meat color values (Table 3). Correlation coefficients between raw and thaw L*, a*, and b* values were in the range of 0.76 to 0.79, and were higher than those between raw-cook, which ranged from 0.50 to 0.70, and thaw-cook, with values between 0.44 and 0.63. Those results are in accordance with the work of Fletcher et al. (2000). Weight. Weights of fillets after removal from the carcass as well as after thawing and cooking were similar, regardless of their initial L* value (Table 4). Thawing losses from the frozen state ( 10%) and cooking losses from thawed state ( 26%) were extensive, but influence of their fresh L* value was small and not statistically significant until both were combined (total loss). Thus, total weight loss from frozen to cooked state was higher for the pale fillets than for the dark fillets. Pearson s correlation coefficient (r = 0.27, P 0.01) confirmed that total losses increased with fresh L* values. Thawed fillet L* value was also correlated to total losses (r = 0.214, P 0.05), and other positive correlations were observed between b* value of thawed fillets and cook loss (r = 0.253, P 0.01) and total loss (r = 0.248, P 0.01). Correlations between fresh L* values and water losses during storage and cooking of broiler breast fillets have been well established (Allen et al., 1998; Le Bihan-Duval et al., 1999; Polidori et al., 2000; Woelfel et al., 2002). No other reports have been found that show correlations between weight losses and b* value. In practice, considerable variation exists for thawing and cooking losses, and values from 1.7 to 18.5% and from 12 to 39%, respectively, are possible, depending on the conditions of the process (Khan and Van den Berg, 1967; Crigler and Dawson, 1968; Cunningham and Lee, 1975; Prusa and Lonergan, 1987; Jakubowska et al., 1997; O Neill et al., 1998; Northcutt et al., 2001; Woelfel and Sams, 2001). Lyophilization was performed on cooked fillets to measure residual water content, and no differences could be detected (grand mean = 64.5%, 0.05) between L* value categories. Similarly, the time needed to achieve 80 C internal temperature during cooking was not altered (grand mean = 38.9 min, 0.05), regardless of the initial L* value. These results and those mentioned above confirm the relationship between breast meat L* and its water-holding capacity (measured by the water losses during freezethawing and cooking). In our experiment, total losses were higher for the pale fillets than for the dark ones. Similarly, Allen et al. (1998) and Woelfel et al. (2002) found that drip and cooking losses of broiler breast meat were higher for those fillets that were pale than for those classified as dark. Woelfel and Sams (2001) reported that pale unmarinated fillets had higher expressible moisture and cook loss than normal marinated fillets. In turkeys, Barbut (1993) found a 9% difference in cooking loss between breast fillets having very light (53.6) or very dark (41.1) L* values 24 h postmortem. Dimensions (Thaw Vs. Cooked). The fillet represents a muscle composed entirely of fast glycolytic fibers oriented in one direction, and the extensive losses from cooking are expected to alter its dimensions (Table 5). No differences were observed in thawed fillet dimensions, regardless of L* value. However, cooking significantly reduced length (51 mm, 24.5%) and width (17 mm, 15.2%) independently of L* value, whereas thickness increased in

FREEZE-THAW, COOKING, AND FILLET PALENESS 2097 a manner that was minimal with pale fillets. Presumably, increased preparation losses associated with pale fillets is compensated by reduced thickness given equivalent moisture content. In fact, positive correlations were found between variation in width and cooking and total losses (r = 0.240 and r = 0.296, respectively, P 0.05) and thickness and cooking and total losses (r = 0.231 and r = 0.222, respectively, P 0.05). Regarding the magnitude of the changes, very little data is available. Lyon et al. (1997) observed a 22% reduction in length and 15% reduction in width of broiler breast meat that had been immersed in water at 85 C for 30 min, and Papa and Lyon (1989) noted a reduction of 24.3% in length after oven cooking. Neither report mentioned change in thickness. In conclusion, variation in broiler breast meat color 48 h postmortem can be used as an indicator of differences in meat quality properties. The fillets having the highest L* values 48 h postmortem within a population suffer higher total water losses associated with thawing and cooking compared with the dark ones. However, the change in dimensions seems not to be affected by the initial color of the sample, with the exception of thickness. Therefore, implementing in-line measurement of fillet L* may be advantageous to sort the fillets according to appropriate use. REFERENCES Abeni, F., and G. Bergoglio. 2001. Characterization of different strains of broiler chicken by carcass measurements, chemical and physical parameters and NIRS on breast muscle. Meat Sci. 57:133 137. Allen, C. D., D. L. Fletcher, J. K. Northcutt, and S. M. Russell. 1998. The relationship of broiler breast color to meat quality and shelf-life. Poult. Sci. 77:361 366. Allen, C. D., S. M. Russell, and D. L. Fletcher. 1997. The relationship of broiler breast meat color and ph to shelf-life and odor development. Poult. Sci. 76:1042 1046. Barbut, S. 1993. Color measurements for evaluating the pale soft exudative (PSE) occurrence in turkey meat. Food Res. Int. 26:39 43. Barbut, S. 1996. Estimates and detection of the PSE problem in young turkey breast meat. Can. J. Anim. Sci. 76:455 457. Claus, J. R., J. K. Schilling, N. G. Marriott, S. E. Duncan, M. B. Solomon, and H. Wang. 2001. Hydrodynamic shockwave tenderization effects using a cylinder processor on early deboned broiler breasts. Meat Sci. 58:287 292. Crigler, J. C., and L. E. Dawson. 1968. Cell disruption in broiler breast muscle related to freezing time. J. Food Sci. 33:248 250. Cunningham, F. E., and H. W. Lee. 1975. Influence of thawing and cooking procedures on selected characteristics of broilers. Poult. Sci. 54:903 908. Fletcher, D. L. 1999. Broiler breast meat color variation, ph, and texture. Poult. Sci. 78:1323 1327. Fletcher, D. L., M. Qiao, and D. P. Smith. 2000. The relationship of raw broiler breast meat color and ph to cooked meat color and ph. Poult. Sci. 79:784 788. Jakubowska, M., J. Kortz, J. Gardzielewska, and T. Karamucki. 1997. Effect of commercial broiler strains on meat weight losses during refrigeration, thawing and cooking. Pages 212 215 in Proc. of the XIII European Symp. on the Quality of Poultry Meat. Poznan, Poland. Khan, A. W., and L. Van den Berg. 1967. Biochemical and quality changes occurring during freezing of poultry meat. J. Food Sci. 32:148 150. Le Bihan-Duval, E., N. Millet, and H. Remignon. 1999. Broiler meat quality: Effect of selection for increased carcass quality and estimates of genetic parameters. Poult. Sci. 78:822 826. Liu, C.-H., and R. Van Laack. 1998. Characteristics of pale, soft, exudative (PSE) broiler breast muscle. Page 15 in Book of abstracts of the IFT s 1998 Annual Meeting, Atlanta, GA. Liu, Y., and Y. R. Chen. 2001. Two-dimensional visible/nearinfrared correlation spectroscopy study of thawing behavior of frozen chicken meats without exposure to air. Meat Sci. 57:299 310. Lyon, B. G., C. E. Lyon, D. P. Smith, and E. M. Savage. 2003. The effects of diet and feed withdrawal times on the sensory descriptive profile and shear values of broiler breast meat. Page: 49 in International Poultry Scientific Forum. Southern Poultry Science Society, Atlanta, GA. Lyon, C. E., S. F. Bilgili, and J. A. Dickens. 1997. Effects of chilling time and belt flattening on physical characteristics, yield, and tenderness of broiler breasts. J. Appl. Poult. Res. 6:39 47. Lyon, C. E., B. G. Lyon, and J. A. Dickens. 1998. Effects of carcass stimulation, deboning time, and marination on color and texture of broiler breast meat. J. Appl. Poult. Res. 7:53 60. Lyon, C. E., W. E. Townsend, and R. L. Wilson. 1976. Objective color values of non-frozen and frozen broiler breasts and thighs. Poult. Sci. 55:1307 1312. McCurdy, R. D., S. Barbut, and M. Quinton. 1996. Seasonal effect on pale soft exudative (PSE) occurrence in young turkey breast meat. Food Res. Int. 29:363 366. Northcutt, J. K., R. J. Buhr, L. L. Young, C. E. Lyon, and G. O. Ware. 2001. Influence of age and postchill carcass aging duration on chicken breast fillet quality. Poult. Sci. 80:808 812. O Neill, L. M., K. Galvin, P. A. Morrissey, and D. J. Buckley. 1998. Comparison of effects of dietary olive oil, tallow and vitamin E on the quality of broiler meat and meat products. Br. Poult. Sci. 39:365 371. Owens, C. M., E. M. Hirschler, S. R. McKee, R. Martinez-Dawson, and A. R. Sams. 2000. The characterization and incidence of pale, soft, exudative turkey meat in a commercial plant. Poult. Sci. 79:553 558. Papa, C. M., and C. E. Lyon. 1989. Shortening of the pectoralis muscle and meat tenderness of broiler-chickens. Poult. Sci. 68:663 669. Petracci, M., and D. L. Fletcher. 2002. Broiler skin and meat color changes during storage. Poult. Sci. 81:1589 1597. Polidori, P., M. T. Marinucci, F. Fantuz, C. Renieri, and F. Polidori. 2000. Pale, soft and exudative (PSE) meat in broiler chickens: characteristics and assessment methods. Industrie Alimentari 39:327 331. Prusa, K. J., and M. M. Lonergan. 1987. Cholesterol content of broiler breast fillets heated with and without the skin in convection and conventional ovens. Poult. Sci. 66:990 994. Qiao, M., D. L. Fletcher, J. K. Northcutt, and D. P. Smith. 2002. The relationship between raw broiler breast meat color and composition. Poult. Sci. 81:422 427. SAS Institute. 2001. SAS version 8.02. SAS Institute Inc., Cary, NC. Wilkins, L. J., S. N. Brown, A. J. Phillips, and P. D. Warriss. 2000. Variation in the colour of broiler breast fillets in the UK. Br. Poult. Sci. 41:308 312. Woelfel, R. L., and A. R. Sams. 2001. Marination performance of pale broiler breast meat. Poult. Sci. 80:1519 1522. Woelfel, R. L., C. M. Owens, E. M. Hirschler, R. Martinez-Dawson, and A. R. Sams. 2002. The characterization and incidence of pale, soft, and exudative broiler meat in a commercial processing plant. Poult. Sci. 81:579 584. Yang, C. C., and T. C. Chen. 1993. Effects of refrigerated storage, ph adjustment, and marinade on color of raw and microwave cooked chicken meat. Poult. Sci. 72:355 362.