Properties of bone from Catfish heads and frames (2024)

2.1. Catfish frames

Catfish frames were acquired from a commercial Catfish processing facility in Mississippi and held in a −20°C freezer untilled thawed at 4°C. Frames also had some fins and ribs attached. Viscera consisted of the stomach, intestine, internal fat pad, some liver and testes, and egg when present. For the control treatment, 740g of Catfish frame was weighed and dried in a Cyclone convection oven (model GDCO‐E, Baker's Pride Oven Company, Cheyenne, WY) for 24hr at 66°C. After drying, Catfish frames were ground using a Waring Pro meat grinder (model WPG200SA, Waring Commercial, Torrington, CT) fitted with a plate having 0.25‐inch‐diameter holes. Catfish frames were then freeze milled using a Spex Sample Prep freezer mill (model 6870D, Spex Sample Prep, Metuchen, NJ) and stored in a −70°C freezer until needed. All treatments were replicated three times.

The pressure wash treatment used 740g of Catfish frames which were placed in boiling water in a Groen Steam Kettle (Model TDA/1‐40) for 20min. Catfish frames were then removed from the Groen Kettle and placed on a drying rack and pressure washed (Ryobi, model RY14122, Techtronic Industries, Anderson, SC) for 2min per side. After pressure washing these frames, they were oven dried for 24hr at 66°C, freeze milled, and stored at −20°C.

Catfish frame enzyme treatments consisted of 740g of Catfish frames grounded with a Hobart grinder with a plate having 0.25‐inch holes. After grinding, approximately 740g of deionized water was added to the ground frames and heated to 50°C for 15min. After equilibration, 355µl of Alcalase (Protease from Bacillus lichenformis, 2.4U/g, Sigma Chemical, St. Louis) was added and the mixture stirred with an overhead propeller for 50°C for 30 or 60min. The enzyme was inactivated by heating to 95°C for 20min. The mixture was then permitted to cool down to room temperature, and the bone was separated from the hydrolysate by filtering through a 35 mesh stainless steel wired screen. It was washed three times with DI water, dried in a Cyclone convection oven at 66°C, and freeze milled.

2.2. Catfish heads

Catfish heads were acquired from a commercial Catfish processing facility in Mississippi. Heads were held in a −20°C freezer until thawed at 4°C. Catfish heads contained the heart and some of the liver. Heads were ground (model 7548, The Biro Mfg. Co., Marblehead, OH) first through a plate with 1‐inch and then 0.5‐inch‐diameter holes. Ground Catfish head material was separated into 300g portions, placed in Ziploc bags and held in a −70°C freezer (So Low Environmental, Cincinnati, OH). All enzyme treatments were performed in triplicate using 300g of semithawed Catfish head that was mixed with 300g of deionized water and hom*ogenized in a Waring Blender (model 38BL61) for 2min. For all treatments, the Catfish head hom*ogenate was preheated to 50°C using a microwave (model NN‐SN936B, Panasonic Appliances, Shanghai, CH). After attaining the 50°C temperature, 355µl of Alcalase (2.4U/g) was added to the Catfish head hom*ogenate and placed in a shaker (Labline Instruments Inc., Melrose Park, IL) set at 110rpm for 0, 5, and 30min at 50°C. At the appropriate time, the hydrolyzed samples were put into a microwave and rapidly brought to 95°C and then placed on a hot plate to maintain a temperature of 95°C for 20min to inactivate Alcalase.

All nonenzyme‐treated samples were subjected to the same procedures except no Alcalase enzyme was added to the hom*ogenates. After the 95°C treatment, hom*ogenates were filtered through a 35 mesh stainless steel sieve screen to separate the bone fragment material from hydrolysate liquids. The separated bone fragments were washed with 200g of deionized water, and the bone fragments were dried in a Cyclone electric convection oven for 24hr at 66°C. After drying, bone samples were weighed and freeze milled. All samples were stored in a −70°C freezer until further analyzed.

2.3. Proximate analysis

Moisture and ash contents were determined using AOAC methods #952.08 and #938.08, respectively (AOAC, ¹⁹⁹⁰). Nitrogen content was accessed by pyrolysis with a LECO FP‐2000 nitrogen analyzer (LECO Co., St. Joseph, MO). Protein content was calculated as 6.25 times % N. Total lipid content was determined gravimetrically by the method of Dodds, McCoy, Geldenhuys, Rea, and Kennish (²⁰⁰⁴) after extraction with an Accelerated Solvent Extractor (ASE; ASE 250, Dionex, Sunnyvale, CA) using methylene chloride. Solvent was removed under a N₂ gas stream at 40°C using a TurboVap LV (Caliper Life Sciences) in preweighed vials. The remaining traces of solvent were removed under vacuum until constant weight was achieved.

Conversion of wet weight proximate percentages to a dry weight percentage was accomplished in Excel. A new set of proximate numbers equal to the wet weight values was created, and a third set of proximate numbers was created equal to the second set multiplied by the ratio of the sum of the original wet weight proximate values to the sum of the second set. Then, the Excel solver function (data tab) was used to change the moisture value of the second set so that the moisture value of the third set equaled the value needed (solver parameters used an objective set so the moisture value of the third set of values was equal to the value being used as the dry weight moisture percentage; set to the moisture value of the treated sample). The results in the second set values were equivalent to the weight of the original sample after drying (only a change in the moisture value). The results in the third set would be the proximate values of the “dry” sample.

Calculations of percent loss of lipid or protein content resulting from mechanical or enzymatic treatments were performed in a similar manner to the conversion of wet weight proximate values to dry weight proximate values above. The dry weight raw proximate values (as calculated above) were converted to the treated (mechanical or enzymatic) proximate values through an intermediate set of values (set 2, as described above). Solver parameters were started with an objective set to the ash value in the third set of values, to be equal to the proximate ash value of the treated sample; the moisture value of the third set was restricted to the proximate moisture value of the treated sample; the ash value of the third set was restricted to values less than or equal to the beginning “dried” sample. The solver function was repeated for lipid and protein, and all three repeated until all four numbers in the third set of values equaled the treated proximate values (the sum of differences, between set 3 and treated values, was <0.01). The percent loss of lipid or protein was then calculated by dividing the difference between the “dried” raw sample proximate value and the optimized second set value, divided by the dried raw sample proximate value, times 100. If the solver function was restricted so the third set of values are kept within 5% of the treated proximate values, it produced an optimized set with fewer repetitions.

2.4. Amino acid analysis

Amino acid profiles were determined by the AAA Service Laboratory Inc. (Boring, OR). Samples were hydrolyzed with 6M HCl and 2% phenol at 110°C for 22hr. Amino acids were quantified using the Beckman 6300 amino acid analyzer (Beckman Coulter, Brea, CA) with postcolumn ninhydrin derivatization. To minimize, methionine losses oxygen was removed by evacuation of the hydrolysis tubes that contained samples and acid for 10min prior to putting them in the hydrolysis oven. Cysteine and tryptophan were not determined as separate hydrolysis procedures are required, which increases analysis cost. Two samples of each type of tissue were analyzed for amino acid composition.

2.5. Mineral analysis

Elemental analysis was conducted at the University of Missouri‐Columbia College of Agriculture, Food and Natural Resources, Chemical Laboratory on dried samples sent for analysis. Samples were wet digested according to AOAC official method 968.08‐D (b) with nitric acid and perchloric acid. After dilution, the filtered samples were introduced into an ICP‐OES (AOAC 985.01‐(A,B,D)).

2.6. Statistical analysis

Enterprise Guide, version 5.1 (SAS, Inc., Cary, NC), was used to conduct the analysis of variance (ANOVA) and Tukey's test on the means data derived from this experimental study. p<0.05 was the relevant significance value used on the data accumulated.

3.1. Effect of process treatments on proximate analysis of Catfish frame bone meal

The percent moisture was highest in the untreated raw frames (13.8%) and was significantly different from the treatments in this study (Table ​(Table1).1). Pressure wash, 30‐min Alcalase treatment, and the 60‐min Alcalase treatment yielded comparable moisture content and showed no significant difference among these respective treatments.

Lipid content was also observed to be the largest in the untreated frames (40.7%), and it was significantly different from the treatments. The pressure wash treatment was significantly higher than the enzyme treatments. Alcalase was noted to be quite efficient in removing lipid from the Catfish frames; however, no significant difference was evidenced between the 30‐ and the 60‐min treatments. The reported lipid content of Catfish frame (Bechtel et al., ²⁰¹⁷), when converted to similar moisture content, was 42 percent, similar to the 41 percent in this study. Lipid content of tuna frame (prepared similarly to the Catfish raw frame in Table ​Table1)1) was reported as 11 percent (Abbey, Glover‐Amengor, Atikpo, Atter, & Toppe, ²⁰¹⁷) or 27 percent (Istiqlaal, ²⁰¹⁷), and for Pollock and Cod, values of 4 and 3 percent were reported (after conversion to 4 percent moisture; Bechtel, ²⁰⁰³)—much lower than the 41 percent in these Catfish samples. In comparison with the treated Catfish frame (1.53 percent lipid), cleaned bones (head and frame) from Cod and Blue whiting had a lipid content of 1.14 and 4.91 percent, respectively, but Salmon and Trout had 38.12 and 34 percent, respectively (Toppe et al., ²⁰⁰⁷).

Protein content of the frame bone showed no significance between the untreated frame and the three treatments, and no significance between the three treatments. However, it was observed that the pressure wash treatment expressed more percent protein than the other treatments in this study. Protein content of the raw frame (33 percent) was very similar to that reported for Catfish frame (converted to a moisture content of 13.8 percent) with 34 percent (Bechtel et al., ²⁰¹⁷). Raw Tuna frame is reported to have 29 percent protein (Abbey et al., ²⁰¹⁷) or 9 percent (Istiqlaal, ²⁰¹⁷). Protein content of cleaned bone (head and frame) of Cod and Blue whiting was reported as 36 and 42 percent, respectively, and Salmon and Trout had 29 and 31 percent, respectively (Toppe et al., ²⁰⁰⁷), compared to the 33 percent found for Catfish in this report.

Percent ash for the 30‐min enzyme treatment of Catfish frame bone meal was not determined. However, the 60‐min treatment yielded the highest value of 62.2 percent, significantly higher than the pressure wash treatment. Both treatments were significantly higher than the untreated frame, with 17 percent. Ash content of the raw frame (17 percent) was larger than that reported (Bechtel et al., ²⁰¹⁷) in Catfish frame (11 percent). For Tuna frame, ash content was reported to be 44 percent (Abbey et al., ²⁰¹⁷). The 62 percent ash for the 60‐min enzyme treated was larger than that reported for Cod, Blue whiting, Salmon, or Trout (53, 45, 26, and 27 percent, respectively; Toppe et al., ²⁰⁰⁷).

From the results in Table ​Table1,1, with the decrease in lipids corresponding with no change in protein content and a large increase in ash content, it was calculated that the pressure wash percentages correspond to an approximate loss of 93 percent of original lipid and 67 percent of original protein. This calculates to a 1.4:1 lipid/protein mixture that was removed from the frame bone. For the 60‐min enzyme treatment, the material lost from the frame bone calculates as approximately 99 percent total lipid and 73 percent total protein in a similar 1.4:1 ratio. Therefore, the tissue that was attached to the frame, and removed by the treatments, was preferentially lipids. The resulting bone product was low in lipid content and high in protein and mineral (ash) content, ideal for a nutritional supplement or ingredient, and an especially good choice to counter micronutrient‐deficient diets of those who suffer with obesity. Dietary intakes of calcium and iron have been found to below the recommended dietary allowance for individuals preparing for bariatric surgery (Frame‐Peterson, Megill, Carobrese, & Schweitzer, ²⁰¹⁷).

3.2. Effect of process treatments on proximate analysis of Catfish head bone meal

Since the 30 and 60min enzyme treatments of Catfish frames were not significantly different, for the Catfish head treatment, shorter time periods were examined. Moisture content of both the 5‐ and 30‐min enzyme treatments of the Catfish head bone were significantly larger than the 0‐min enzyme treatment (Table ​(Table2);2); however, only the 30‐min treatment was larger than the untreated head bone.

With 29.2 percent lipid, the untreated head was significantly larger than the three enzyme treatments. The 5‐min enzyme treatment was significantly smaller than the 0‐min treatment, and the 30‐min treatment was significantly smaller than the 5‐min treatment, showing a progressive loss of lipid during enzyme treatment. The lipid content of head bone from Pollock, Cod, and Salmon was reported as 6, 4, and 11 percent, respectively (after conversion to 4% moisture; Bechtel, ²⁰⁰³). These values differ greatly from the Catfish raw head, and more closely match the enzyme treated values seen in Table ​Table22.

Protein content of the untreated head bone and the 0‐min enzyme treatment were not significantly different, but they were both significantly larger than the 5‐min or 30‐min treatments, which were not significantly different. Protein content of Pollock, Cod, and Salmon head bone was reported as 70, 75, and 47 percent, respectively (after conversion to 4% moisture; Bechtel, ²⁰⁰³).

3.3. Effect of process treatments on mineral content of Catfish frame bone meal

On average, the Catfish frame mineral content, seen in Table ​Table3,3, shows the 30‐ and 60‐min enzyme treatments to be significantly larger than the pressure wash or untreated frame bone. However, sodium was an exception, with the pressure wash being significantly larger than either enzyme treatment. Potassium in the untreated frame bone was significantly larger than any of the three treatments. For iron, there was not a significant difference between untreated or treated samples. The 60‐min enzyme treated frame bone contained 25 percent calcium and a 2.2 ratio for Ca:P. Raw Pollock and Cod frame bone was reported to contain 5.9 and 6.0 percent calcium, respectively (Bechtel & Johnson, ²⁰⁰⁴), similar to the Raw Catfish Frame value of 6.3 percent (Table ​(Table33).

3.4. Effect of process treatments on mineral content of Catfish head bone meal

The general trend for mineral content of head bone for the three enzyme treatments was a significant increase with increasing treatment time from 0 to 5 to 30min (Table ​(Table4).4). However, sodium, potassium, and nickel did not show a difference between treatment times, and iron showed a reverse direction, with significant decreases with enzyme treatment time. The 30‐min enzyme treated head bone contained 20 percent calcium and a 2.1 ratio for Ca:P. Raw Pollock and Cod head bone was reported to contain 6.6 and 5.6 percent calcium, respectively (Bechtel & Johnson, ²⁰⁰⁴), but the 0‐min enzyme treatment in Table ​Table44 is different from raw Catfish head bone since it was heated in water and filtered through a mesh.

3.5. Effect of process treatments on amino acid content of Catfish frame bone meal

As seen in Table ​Table5,5, essential amino acids (EAA) showed a significant decrease from untreated frame bone (33.2%) to pressure wash (21.5%) to enzyme treatments (19.4%–19.7%). Hydroxyproline (Hyp) concentration showed a reverse trend with a significant increase from the untreated frame (3.9%) to pressure wash (7.1%) to enzyme treatments (7.7%–8.0%), with both enzyme treatments having no significant difference. Proline also significantly increases from the untreated frame to the three treatments. Almost all other amino acids show a significant decrease from the untreated frame to the three treatments. The most abundant amino acid in the 60‐min enzyme treatment was glycine, alanine, proline, glutamic acid, and hydroxyproline on a mole basis. The high content of hydroxyproline is due in part to the connective tissue in bone. The hydroxyproline content of Cod, Salmon, and Trout bones was reported as 5, 6, and 6 percent (Toppe et al., ²⁰⁰⁷), compared to 8 percent for the enzyme treatments and 7 percent for the pressure wash in this study. The relative amino acid content of the treated Catfish frame bone is very similar to collagen, and therefore if hydrolyzed, may be useful as a dietary supplement or functional food or beverage for the benefit of joint and bone health or enhancement of skin health.

3.6. Effect of process treatments on amino acid content of Catfish head bone meal

Table ​Table66 shows the EAA significantly decreases from 0‐min enzyme treatment (30.3%) to 5min (28.3%) to 30min (25.1%) in Catfish head bone meal. Hydroxyproline and proline show a reversed trend with increasing concentrations with longer enzyme treatment, from 4.4 percent to 6.1 percent.

• Abbey, L., Glover‐Amengor, M., Atikpo, M., Atter, A., & Toppe, J. (2017). Nutrient content of fish powder from low value fish and fish byproducts. Food Science & Nutrition, 5(3), 374–379. 10.1002/fsn3.402 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• AOAC. (1990). Official methods of analysis (15th edn). Arlington,VA: Association of Official Analytical Chemists. [Google Scholar]
• Bechtel, P. J. (2003). Properties of different fish processing by‐products from pollock, cod and salmon. Journal of Food Processing and Preservation, 27, 101–116. 10.1111/j.1745-4549.2003.tb00505.x [CrossRef] [Google Scholar]
• Bechtel, P., Bland, J., Bett‐Garber, K., Grimm, C., Brashear, S., Lloyd, S., … Lea, J. (2017). Chemical and nutritional properties of channel and hybrid catfish byproducts. Food Science & Nutrition, 5(5), 981–988. 10.1002/fsn3.483 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Bechtel, P. J., & Johnson, R. (2004). Nutritional properties of pollock, cod and salmon processing byproducts. Journal of Aquatic Food Product Technology, 13(2), 125–142. 10.1300/j030v13n02_11 [CrossRef] [Google Scholar]
• Bosworth, B. G., Wolters, W. R., Silva, J. L., Chamul, R. S., & Park, S. (2004). Comparison of production, meat yield, and meat quality traits of NWAC103 line channel catfish, Norris line channel catfish, and female channel catfish X male blue catfish F₁ hybrids. North American Journal of Aquaculture, 66, 177–183. 10.1577/a03-032.1 [CrossRef] [Google Scholar]
• Crapo, C., & Bechtel, P. J. (2003). Utilization of Alaska’s seafood processing byproducts In Bechtel P. J. (Ed.), Advances in seafood byproducts: 2002 Conference proceedings (pp. 105–119). Anchorage, AK: Alaska Sea Grant, University of Alaska Fairbanks. [Google Scholar]
• Davenport, M. P., & Kristinsson, H. G. (2011). Channel catfish (Ictalurus punctatus) muscle protein isolate performance processed under different acid and alkali pH values. Journal of Food Science, 76, E240–E247. 10.1111/j.1750-3841.2011.02083.x [PubMed] [CrossRef] [Google Scholar]
• Dodds, E. D., McCoy, M. R., Geldenhuys, A., Rea, L. D., & Kennish, J. M. (2004). Microscale recovery of total lipids from fish tissue by accelerated solvent extraction. Journal of the American Oil Chemists' Society, 81(9), 835–840. 10.1007/s11746-004-0988-2 [CrossRef] [Google Scholar]
• Frame‐Peterson, L. A., Megill, R. D., Carobrese, S., & Schweitzer, M. (2017). Nutrient deficiencies are common prior to bariatric surgery. Nutrition in Clinical Practice, 32(4), 463–469. 10.1177/0884533617712701 [PubMed] [CrossRef] [Google Scholar]
• Hanson, T., & Sites, D. (2012). 2011 U.S. Catfish Database. Retrieved from https://www.aces.edu/dept/fisheries/aquaculture/catfish-database/catfish-2011.php. [last accessed 26 January 2019].
• Hoke, M. E., Jahncke, M. L., Silva, J. L., Hernsberber, R. S., & Suriyaphan, O. (2000). Stability of washed frozen mince from channel catfish frames. Journal of Food Science, 65, 1083–1086. 10.1111/j.1365-2621.2000.tb09422.x [CrossRef] [Google Scholar]
• Istiqlaal, S. (2017). Proximate levels of bone bluefin tuna fish as gelatinization by product. IOSR Journal of Environmental Science, Toxicology and Food Technology, 11(4), 12–17. 10.9790/2402-1104021217 [CrossRef] [Google Scholar]
• Jiang, M., Shaoyang, L., Du, X., & Wang, Y. (2010). Physical properties and internal microstructures of films from catfish skin gelatin and triacetin mixtures. Food Hydrocolloids, 24, 105–110. 10.1016/j.foodhyd.2009.08.011 [CrossRef] [Google Scholar]
• Lovell, R. T. (1973). Put catfish offal to work for you. Fish Farming Industries October/ November, 22–24.
• Nguyen, T. T., Lindberg, J. E., & Ogle, B. (2007). Survey of the production, processing and nutritive value of catfish by‐product meals in the Mekong Delta of Vietnam. Livestock Research for Rural Development, 19(9), article #124. [Google Scholar]
• Ren, X., Ma, L., Chu, J., Wang, Y., Zhuang, Y., Zhang, S., & Yang, H. (2012). Optimization of enzymatic hydrolysis of channel catfish bones for preparing antimicrobial agents. Journal of Aquatic Food Product Technology, 21(2), 99–110. 10.1080/10498850.2011.586136 [CrossRef] [Google Scholar]
• Sathivel, S., Prinyawiwatkul, W., Grimm, C. C., King, J. M., & Lloyd, S. W. (2002). Fatty acid composition of crude oil recovered from catfish viscera. Journal of the American Oil Chemists' Society, 79, 989–992. 10.1007/s11746-002-0592-5 [CrossRef] [Google Scholar]
• Sathivel, S., Yin, H., Bechtel, P. J., & King, J. M. (2009). Physical and nutritional properties of spray dried protein powders from catfish roe and its application in an emulsion system. Journal of Food Engineering, 95, 76–81. 10.1016/j.jfoodeng.2009.04.011 [CrossRef] [Google Scholar]
• Sathivel, S., Yin, H., Prinyawiwatkul, W., & King, J. M. (2009). Comparison of chemical and physical properties of catfish oils prepared from different extracting processes. Journal of Food Science, 74, E70–E76. 10.1111/j.1750-3841.2009.01050.x [PubMed] [CrossRef] [Google Scholar]
• Silva, J., & Dean, S. (2001). Processed Catfish: Product forms, packaging, yields and product mix. SRAC Publication No. 184. MSU Food and Fiber Center.
• Techochatchawal, K., Therdthai, N., & Khotavivattana, S. (2009). Development of calcium supplement from the bone of Nile tilapia (Tilapia nilotica). Asian Journal of Food & Agro‐industry, 2, 539–546. [Google Scholar]
• Theodore, A. E., Raghavan, S., & Kristinsson, H. G. (2008). Antioxidative activity of protein hydrolysates prepared from alkaline‐aided channel catfish protein Isolates. Journal of Agricultural and Food Chemistry, 56, 7459–7466. 10.1021/jf800185f [PubMed] [CrossRef] [Google Scholar]
• Toppe, J., Albrektsen, S., Hope, B., & Aksnes, A. (2007). Chemical compostion, mineral content and amino acid and lipid profiles in bones from various fish species. Comparative Biochemistry and Physiology B, 146, 395–401. 10.1016/j.cbpb.2006.11.020 [PubMed] [CrossRef] [Google Scholar]
• Venkatesan, J., Lowe, B., Manivasagan, P., Kang, K., Chalisserry, E., Anil, S., … Kim, S. (2015). Isolation and characterization of nano‐hydroxyapatite from salmon fish bone. Materials, 8(8), 5426–5439. 10.3390/ma8085253 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Wiles, J. L., Green, B. W., & Bryant, R. (2004). Texture profile analysis of a minced catfish product. Journal of Texture Studies, 35, 325–337. 10.1111/j.1745-4603.2004.tb00841.x [CrossRef] [Google Scholar]
• Woodruff, V. C. (1984). Processing trends in the U.S. catfish industry. Unpublished Paper. Mississippi State U., MS.
• Wu, G., Zhang, M., Wanga, Y., Mothibe, K., & Chen, W. (2012). Production of silver carp bone powder using superfine grinding technology: Suitable production parameters and its properties. Journal of Food Engineering, 109, 730–735. 10.1016/j.jfoodeng.2011.11.013 [CrossRef] [Google Scholar]
• Yang, H., Wang, Y., Jiang, M., Oh, J., Herring, J., & Zhou, P. (2007). 2‐Step optimization of the extraction and subsequent physical properties of channel catfish (Ictalurus punctatus) skin gelatin. Journal of Food Science, 72, C188–C195. 10.1111/j.1750-3841.2007.00319.x [PubMed] [CrossRef] [Google Scholar]
• Yin, H., Pu, J., Wan, Y., Xiang, B., Bechtel, P. J., & Sathivel, S. (2010). Rheological and functional properties of Catfish skin protein hydrolysates. Journal of Food Science, 75, E11–E17. 10.1111/j.1750-3841.2009.01385.x [PubMed] [CrossRef] [Google Scholar]
• Yin, H., Wan, Y., Pu, J., Bechtel, P. J., & Sathivel, S. (2011). Functional and rheological properties of protein fractions of channel catfish (Ictalurus punctatus) and their effects in an emulsion system. Journal of Food Science, 76, E283–E290. 10.1111/j.1750-3841.2011.02057.x [PubMed] [CrossRef] [Google Scholar]
• Yin, T., Du, H., Xhang, J., & Xiong, S. (2016). Preparation and characterization of ultrafine fish bone powder. Journal of Aquatic Food Product Technology, 25(7), 1045–1055. 10.1080/10498850.2015.1010128 [CrossRef] [Google Scholar]
• Yin, T., Park, J., & Xiong, S. (2015). Physicochemical properties of nano fish bone prepared by wet media milling. LWT ‐ Food Science and Technology, 64(1), 367–373. 10.1016/j.lwt.2015.06.007 [CrossRef] [Google Scholar]