Predicting Objective and Subjective Pork Belly Firmness Using Fatty Acid Indices

Justice B. Dorleku; Tawanda Tayengwa; David Rolland; Manuel Juárez; Justice B. Dorleku; Tawanda Tayengwa; David Rolland; Manuel Juárez

doi:10.22175/mmb.21371

Introduction

Pork belly has become an important primal cut due to increasing domestic and international consumer demand. The main quality trait impacting its value is pork belly firmness, with softer pork bellies being undesirable due to challenges in further processing and sliceability, both in raw and processed bellies (Soladoye et al., 2015). Pork belly firmness is a multifactorial quality trait, influenced by dimensional traits such as thickness and length, as well as chemical characteristics, including fatty acid (FA) composition (Soladoye et al., 2017). The concentration of long-chain unsaturated fatty acids (UFAs) influences fat firmness, and the pork sector usually measures it using the iodine value (IV). Researchers and industry have employed various methods to measure or predict IV, including calculations derived from individual FA gas chromatography (GC) analysis. The American Oil Chemists’ Society (AOCS) has published an official method to calculate IV based on individual FAs (AOCS, 2009). Other authors have proposed alternative methods for calculating iodine value (IV), such as including all UFAs or using different formulas. For instance, Lo Fiego et al. (2016) recommended 2 equations (Back 2 and Step 2) that enhanced the estimation of IV in the subcutaneous fat of Italian heavy pigs. In industry, both FA indices and individual FAs are sometimes considered together. For example, the production of Italian Protected Designation of Origin hams requires fat IV not to exceed 70, with a C18:2 content under 15% (MIPAF, 2007). This can also be replicated for commercial market-weight pigs and incorporated into inline near infrared probes.

The relative contribution of IV to belly firmness seems to be influenced by environmental factors, such as the measuring method and belly temperature. Whitney et al. (2006), using the traditional bar bend method, reported that IV explained 14% of the variability in belly firmness. Wei et al. (2023) evaluated belly firmness using an automated method at 3 different temperatures (−1.5°C, 2°C, 4°C) and explained a range of 18–64% in belly firmness variability. Monziols et al. (2007) reported that the concentrations of saturated (SFA), monounsaturated (MUFA), and polyunsaturated fatty acids (PUFA) did not differ significantly between the subcutaneous fat of shoulder and belly primal cuts. Consequently, previous studies have reported that shoulder fat can be a good predictor of belly firmness (Lam et al., 2021). This is the area of the carcass where IV is measured in commercial abattoirs using technologies such as an inline near infrared probe (NitFom^™). Most of these studies and methodologies use the AOCS FA formula as the reference for IV calculation. Due to the increasing importance of belly firmness in the pork industry, it is necessary to determine if alternative indices can provide additional information to commercial packers and researchers. Furthermore, PUFA profiles are known to provide a direct and more accurate measure of unsaturation, which may be considered superior to traditional IV models. Therefore, this study aimed to compare multiple FA-based indices for their ability to predict pork belly firmness, hypothesizing that PUFA profiles, particularly omega-6 (n-6) content, would outperform conventional IV measures.

Materials and Methods

All experimental procedures were approved by the Agriculture and Agri-Food Canada Lacombe Research and Development Centre’s (AAFC-LRDC, AB, Canada) Animal Care Committee (#201807).

Animal background and sampling procedure

A total of 616 commercially sourced pigs (from Large White × Landrace sows bred to Duroc boars; Genesus Genetic Technology, MB, Canada) were managed under the same commercial feeding program, receiving identical diet formulations adjusted according to standard phase feeding protocols. The pigs were slaughtered at the AAFC-LRDC federally inspected abattoir. Following commercial slaughtering procedures, carcass sides were railed into a 2°C drip cooler room with wind speeds of 0.5 m/s. After 24 h postmortem, left side carcasses were fabricated into primal cuts (picnic, butt, loin, belly, and ham) according to the International Meat Purchase Specifications (IMPS, 2014) for pork, and primal cuts were individually weighed. In alignment with industry practices that estimate IV at the outer subcutaneous fat layer of the pork shoulder, a 5-g sample of shoulder subcutaneous fat was taken from each carcass and stored at −80°C until further FA analysis.

FA composition analysis: GC

Each sample was thawed, and approximately 2 mg was weighed into a 2 mL vial and vacuum dried overnight. Subsequently, 35 microliters (μL) of hexane were added and mixed thoroughly. Derivatization was initiated by adding 140 μL of 0.5 N sodium methoxide in methanol, followed by mixing and incubation in a heat block at 50°C for 10 min. Samples were then cooled to room temperature for 5 min. Next, 210 μL of 5% methanolic hydrochloric acid was added, mixed, and incubated at 80°C for 10 min, then cooled to room temperature for another 10 min. Following derivatization, 1 mL of 19:0 methyl ester in hexane (0.1 mg/mL) was added as an internal standard. Subsequently, 200 μL of distilled water was added, mixed vigorously, and the samples were centrifuged at 1000 × g for 5 min to facilitate phase separation. An aliquot of the upper phase was then analyzed for fatty acid methyl esters (FAMEs) using GC.

A Bruker Scion 436 gas chromatograph with CP-8400 autosampler, 1079 injector, and flame ionization detector equipped with a Phenomenex Zebron ZB-WAXplus capillary GC column (30 m length × 0.25 mm ID, 0.25 μm film thickness) was used. Samples (1 μL) were injected into the injector held at 250°C in split injection mode with a split ratio of 20:1. The carrier gas was hydrogen in constant flow mode at a flow rate of 1 mL/min. The column oven temperature program had an initial temperature of 175°C held for 15 min, then ramped to 250°C at 5°C/min and held for 2 min for a total run time of 32 min. The detector temperature was held at 250°C, and the detector gas flows were 29 mL/min make-up gas (air), 30 mL/min hydrogen, and 300 mL/min air. Scion Instruments Compass CDS Version 4.0.1 software was used to control the gas chromatograph and capture and process chromatographic data. NU-CHEK-PREP GLC reference standard 897 was used for the identification of FAME.

IV equations

Based on the analyzed FA composition, the IV of fat was calculated through the following equations, where square brackets represent the proportion of an individual FA (% of total FA):

AOCS IV = [16 : 1] × 0.95 + [18 : 1] × 0.86 + [18 : 2] × 1.732 + [18 : 3] × 2.616 + [20 : 1] × 0.785 + [22 : 1] × 0.723 Proposed by AOCS (2009) .

All UFA IV = [16 : 1] × 0.950 + [17 : 1] × 0.903 + [18 : 1] × 0.860 + [18 : 2] × 1.732 + [18 : 3] × 2.615 + [20 : 1] × 0.785 + [20 : 2] × 1.580 + [20 : 3] × 2.386 + [20 : 4] × 3.201 + [22 : 4] × 2.941 + [22 : 5] × 3.697 Proposed by Pétursson (2002) and Lo Fiego et al. (2016) .

Back 2 IV = 85.703 + [14 : 0] × 2.740 − [16 : 0] × 1.085 − [18 : 0] × 0.710 + [18 : 2] × 0.986 Proposed by Lo Fiego et al . (2016) .

Step 2 IV = 16.106 + [14 : 0] × 4.611 − [16 : 0] × 0.491 + [18 : 1] × 0.773 + [18 : 2] × 1.728 + [20 : 4] × 3.570 Proposed by Lo Fiego et al . (2016) .

In addition, total saturated fatty acids (SFA), total monounsaturated fatty acids (MUFA), total PUFA, omega-3 (n-3), and omega-6 (n-6) fatty acids were also calculated.

Belly firmness measurement

Bellies (IMPS #408B; skin-on, untrimmed, bone-in) were handled carefully to avoid flexing them before their assessments in a cooler set at 2°C. Belly firmness was objectively assessed by measuring the belly bend angle (°), following the procedure described by Uttaro et al. (2024) and using the Belly Bender Version 2 (BBV2) with the bending point for measurements standardized at 24 cm. Following bending, all hard and soft bones/cartilages were removed from bellies in a single sheet, and the ribbed bellies returned to the 2°C cooler to lie flat for 1–2 h. Belly firmness was then subjectively assessed by an experienced evaluator following the method described by Uttaro et al. (2020), where belly floppiness was described as “flabbiness/rollover as determined by moving the caudal end towards the cranial end, skin side out,” and evaluated from 0 (Rolls up closely like a sleeping bag and does not unroll) to 5 (Stays flat; may make a slight bend).

Statistical analysis

Before analysis, data distribution was assessed using PROC UNIVARIATE to confirm normality. After model fitting, residuals were evaluated to ensure that the assumptions of normality and homoscedasticity were met. Potential outliers were examined using studentized residuals (>|3|) and removed only when biologically implausible. Summary statistics were determined using the MEANS procedure in SAS (version 9.4, SAS Inst. Inc., Cary, NC, USA). For the regression analysis of objective and subjective belly firmness against all the FA indices, the PROC REG procedure was employed. All comparisons were preplanned, so no corrections for multiple comparisons were applied. Statistical significance was consistently declared at P < 0.05.

Results and Discussion

Descriptive statistics

Research on testing different FA calculations as predictors of commercial pork belly firmness has emerged as a critical area of inquiry due to its economic and quality implications in the pork industry. The descriptive statistics of the FA indices and pork belly firmness, using objective and subjective measurement techniques, are shown in Table 1. Specifically, objective belly firmness, measured by BBV2, exhibited moderate variability (coefficient of variation, CV = 15.6%) compared to subjective firmness, assessed by floppiness (CV = 45.1%). The BBV2 range showed a widespread variation in firmness across the sample population, and the level of floppiness was similar to the flop score range reported by Harsh et al. (2017). Nevertheless, the variability of both methods can be attributed to the fact that pork belly firmness is influenced by multiple factors, including FA composition, belly thickness, and dimensional traits (Soladoye et al., 2017). Therefore, the authors recommended combining both dimensional and chemical predictors for predicting objective and subjective belly firmness, rather than using IV by itself. Consequently, the indicators of unsaturation in FAs (AOCS IV, All UFA IV, Back 2 IV, and Step 2 IV) remained stable across samples, with low variability (CV = 2.4–5.2%). Moreover, the calculated IVs ranged from 50.7 to 71.4, with an average IV of approximately 60. This average falls below the industry threshold of 74 (Seman et al., 2013), indicating that the fat quality of the bellies is acceptable. Among the FA classes, variation was relatively lower for SFA and MUFA (CV < 6%) than PUFA, including n-3 and n-6 (CV = 13.3–17.2%). The observed variations for the FA classes are comparable to those reported by Lam et al. (2023) for both the calibration and validation sets, specifically for belly fat.

Table 1.

Descriptive statistics of fatty acid indices and related belly traits.

Variable^¹	N	Mean	Std Dev	CV, %	Minimum	Maximum
BBV2 (°)	616	145	22.7	15.6	107	179
Floppiness	515	2.44	1.10	45.1	0.50	5.00
AOCS IV	616	58.9	3.04	5.15	50.7	67.6
All UFA IV	616	61.8	3.24	5.24	53.1	71.4
Back 2 IV	616	61.8	1.47	2.37	58.0	69.6
Step 2 IV	616	59.2	2.94	4.96	51.4	67.4
SFA (%)	616	43.2	2.38	5.51	36.1	50.3
MUFA (%)	616	44.2	2.12	4.79	39.0	51.5
PUFA (%)	616	12.6	1.67	13.3	8.35	18.1
n-3 (%)	616	1.58	0.27	17.2	0.94	2.55
n-6 (%)	616	11.0	1.47	13.3	7.40	15.9

BBV2: belly bender version 2; AOCS IV: Iodine value from equation proposed by the American Oil Chemists’ Society; All UFA IV: Iodine value prediction including all unsaturated fatty acids; Back 2 IV: Regression model using backward elimination procedure (Lo Fiego et al., 2016); Step 2 IV: Regression model using stepwise regression procedure (Lo Fiego et al., 2016); SFA: saturated fatty acids; MUFA: monounsaturated fatty acids; PUFA: polyunsaturated fatty acids; n-3: omega-3; n-6: omega-6; Floppiness score: 0 = very soft, 5 = very hard.

Regression analysis

Iodine value predictions using the AOCS equation (i.e., AOCS IV) and a modified equation that incorporates all unsaturated FAs (i.e., All UFA IV) were found to underestimate IV by approximately 28 and 7%, respectively, for Italian heavy pigs (Lo Fiego et al., 2016). Additionally, the authors reported an overestimation of nearly 8% of samples for IV predictions generated using both backward elimination (i.e., Back 2 IV) and stepwise regression (i.e., Step 2 IV) variable selection methods, when applied to the full set of saturated and unsaturated FAs. Given these discrepancies, it was necessary to reassess the predictive performance of these equations in estimating the firmness of commercial pork bellies. As IV is derived from FA composition, and individual FAs possess distinct rheological and structural properties, it is reasonable that predicted IV accounted for a substantial portion of the variation in belly firmness (Lam et al., 2022).

Table 2 shows the results of the prediction accuracies (R²) of the 4 IV predictive models and various FAs on both BBV2 and floppiness measures of pork belly firmness. Among these predictors, PUFA and n-6 showed the strongest associations with both traits, explaining up to 44% of the variation in BBV2 and 50% of the variation in floppiness. These R² values are significantly higher than most IV equations, except Step 2 IV, suggesting that higher levels of PUFA and n-6 are strongly linked to softer bellies, likely due to their influence on fat fluidity and structural integrity. This aligns with existing knowledge that increased unsaturation in adipose tissue reduces firmness. Studies by Rentfrow et al. (2003) and Soladoye et al. (2017) emphasize the important role of linoleic acid (C18:2n-6), a major component of n-6, a type of PUFA, in determining pork belly softness. Rentfrow et al. (2003) reported that pork fat containing more than 14% of linoleic acid is associated with increased fat softness. Complementing this, Soladoye et al. (2017) also found that linoleic acid had a stronger relationship with belly softness than IV. Overall, the results suggest that PUFA and n-6 may be useful for partial predictions of belly floppiness.

Table 2.

Fatty acid predictors of objective (BBV2) and subjective (floppiness) pork belly firmness.^¹

Parameter	BBV2			Floppiness
Parameter	R²	P Value	F Value	R²	P Value	F Value
AOCS IV	0.35	<0.001	330	0.35	<0.001	276
All UFA IV	0.35	<0.001	334	0.36	<0.001	289
Back 2 IV	0.09	<0.001	62.4	0.19	<0.001	118
Step 2 IV	0.36	<0.001	343	0.35	<0.001	277
SFA	0.17	<0.001	122	0.14	<0.001	83.6
MUFA	0.00	0.122	2.40	0.02	0.001	11.8
PUFA	0.43	<0.001	469	0.50	<0.001	503
n-3	0.23	<0.001	182	0.25	<0.001	167
n-6	0.44	<0.001	478	0.50	<0.001	523

BBV2: belly bender version 2; AOCS IV: Iodine value from equation proposed by the American Oil Chemists’ Society; All UFA IV: Iodine value prediction including all unsaturated fatty acids; Back 2 IV: Regression model using backward elimination procedure (Lo Fiego et al., 2016); Step 2 IV: Regression model using stepwise regression procedure (Lo Fiego et al., 2016); SFA: saturated fatty acids; MUFA: monounsaturated fatty acids; PUFA: polyunsaturated fatty acids; n-3: omega-3; n-6: omega-6; Floppiness score: 0 = very soft, 5 = very hard.

The performance of the IV equations to predict chemically measured IV has been previously evaluated for Italian heavy pigs (Lo Fiego et al., 2016). Consequently, the authors found better results (R² = 0.64–0.81) for All UFA IV, Step 2 IV, and Back 2 IV. In the current study, the 4 IV predictive models, except Back 2 IV, were consistent and moderately strong predictors, explaining 35–36% of the variation for both BBV2 and floppiness (Table 2). In a previous study, 52% of the variation in belly bend angle was explained by the calculated AOCS IV, using samples collected at the belly (Lam et al., 2022). When assessing the contribution of NitFom IV, which is measured at the shoulder, the results align with the present study, showing an R² value of approximately 0.40 (Lam et al., 2022). For Back 2 IV, only 9 and 19% of the variation in BBV2 and floppiness, respectively, were explained. The low R² for Back 2 IV may be due to the exclusion of relevant FAs, potential multicollinearity among included variables, or the model’s inability to capture complex or nonlinear relationships that more accurately reflect the multifactorial nature of pork belly firmness. These results suggest that Back 2 IV is less reliable for predicting belly firmness.

SFA and n-3 exhibited weak predictive ability. Trusell et al. (2011) found a strong and positive correlation between SFA and belly firmness; however, SFA accounted for only 17 and 14% of the variation in BBV2 and floppiness, respectively, in the present study. A smaller but relevant role could be played by n-3, as 23–25% of the variation in firmness traits was explained. Soladoye et al. (2017) found no correlation between MUFA and belly-flop angle or subjective belly softness, consistent with no significant observation in predicting BBV2 and explaining 2% of the variation in floppiness in our study. This suggests that MUFA plays a limited role in determining firmness. This further underscores the dominant role of PUFA, particularly n-6, in determining belly firmness, rather than total unsaturation.

Conclusions

This study found that while traditional IV-based indices offer moderate predictive power, specific FA indices, particularly PUFA and n-6, serve as more robust and biologically relevant predictors of both objective and subjective firmness. These findings highlight the limitations of using IV alone and support a shift toward incorporating targeted FA profiling for more accurate quality assessment. Moreover, the results provide a foundation for developing rapid, online classification tools that integrate FA data, such as PUFA and n-6, thereby enhancing precision in pork grading systems and supporting value-based marketing strategies.

Conflict of Interest

The authors declare no conflicts of interest.

Acknowledgements

Data for this study were generated thanks to the funding provided to 2 larger projects: “Enhancing pork belly quality across the value-chain. (SCAP-ASC-10 Activity 18A),” funded by Swine Innovation Porc within the Swine Cluster 4, and “Evaluating the impact of increased pig harvest weights on carcass composition, product quality and palatability: implications for genomic selection,” funded by Results Driven Agriculture Research (#2024F2553R). Authors are also grateful for the technical support from the meat quality team at the Agriculture and Agri-Food Canada Lacombe Research and Development Centre (Alberta, Canada).

Author Contribution

Justice B. Dorleku: Data curation, Methodology, Investigation, Writing – original draft, Writing – review & editing. Tawanda Tayengwa: Software, Validation, Formal analysis, Data curation, Writing – review & editing. David Rolland: Data curation, Writing – review & editing. Manuel Juárez: Conceptualization, Methodology, Data curation, Supervision, Funding acquisition, Project administration, Writing – review & editing.

Literature Cited

American Oil Chemists’ Society (AOCS). 2009. Official methods and recommended practices of the AOCS. AOCS Press, Champaign, IL.

Harsh, B. N., E. K. Arkfeld, D. A. Mohrhauser, D. A. King, T. L. Wheeler, A. C. Dilger, S. D. Shackelford, and D. D. Boler. 2017. Effect of hot carcass weight on loin, ham, and belly quality from pigs sourced from a commercial processing facility. J. Anim. Sci. 95:4958–4970. doi: https://doi.org/10.2527/jas2017.1674

International Meat Purchase Specifications (IMPS) 2014. Fresh pork series 400 institutional meat purchase specifications. https://www.ams.usda.gov/sites/default/files/media/IMPS_400_Fresh_Pork%5B1%5D.pdf. (Accessed 26 November 2025).https://www.ams.usda.gov/sites/default/files/media/IMPS_400_Fresh_Pork%5B1%5D.pdf

Lam, S., O. P. Soladoye, N. Prieto, B. Uttaro, J. L. Aalhus, I. Larsen, P. Shand, C. Gariépy, and M. Juárez. 2021. Performance of near-infrared spectroscopy in pork shoulder as a predictor for pork belly softness. Can. J. Anim. Sci. 101:386–389. doi: https://doi.org/10.1139/cjas-2020-0049

Lam, S., B. Uttaro, B. M. Bohrer, M. Duarte, and M. Juárez. 2022. Can in-line iodine value predictions (NitFom^™) be used for early classification of pork belly firmness? Foods 11:148. doi: https://doi.org/10.3390/foods11020148

Lam, S., D. Rolland, S. Zawadski, X. Wei, B. Uttaro, and M. Juárez. 2023. Performance of a handheld near-infrared spectroscopy device to predict pork primal belly fat iodine value and loin lean intramuscular fat content. Foods 12:1629. https://doi.rg/10.3390/foods12081629 https://doi.rg/10.3390/foods12081629

Lo Fiego, D. P., G. Minelli, L. A. Volpelli, A. Ulrici, and P. Macchioni. 2016. Calculating the iodine value for Italian heavy pig subcutaneous adipose tissue from fatty acid methyl ester profiles. Meat Sci. 122:132–138. doi: https://doi.org/10.1016/j.meatsci.2016.08.002

MIPAF 2007. Disciplinare di produzione della denominazione d’origine protetta “Prosciutto di San Daniele.” Gazzetta Ufficiale della Repubblica Italiana, supplemento ordinario alla Gazzetta Ufficiale N. 77 del 2 aprile 2007Roma: Istituto Poligrafico e Zecca dello Stato.

Monziols, M., M. Bonneau, A. Davenel, and M. Kouba. 2007. Comparison of the lipid content and fatty acid composition of intermuscular and subcutaneous adipose tissues in pig carcasses. Meat Sci. 76:54–60. doi: https://doi.org/10.1016/j.meatsci.2006.10.013

Pétursson, S. 2002. Clarification and expansion of formulas in AOCS recommended practice Cd 1c-85 for the calculation of iodine value from FA composition. J. Am. Oil Chem. Soc. 79:737–738. doi: https://doi.org/10.1007/s11746-002-0551-1

Rentfrow, G., T. E. Sauber, G. L. Allee, and E. P. Berg. 2003. The influence of diets containing either conventional corn, conventional corn with choice white grease, high oil corn, or high oil high oleic corn on belly/bacon quality. Meat Sci. 64:459–466. doi: https://doi.org/10.1016/S0309-1740(02)00215-2

Seman, D. L., W. N. G. Barron, and M. Matzinger. 2013. Evaluating the ability to measure pork fat quality for the production of commercial bacon. Meat Sci. 94:262–266. doi: https://doi.org/10.1016/j.meatsci.2013.01.009

Soladoye, P. O., P. J. Shand, J. L. Aalhus, C. Gariépy, and M. Juárez. 2015. Review: Pork belly quality, bacon properties and recent consumer trends. Can. J. Anim. Sci. 95:325–340. doi: https://doi.org/10.4141/cjas-2014-121

Soladoye, O. P., B. Uttaro, S. Zawadski, M. E. R. Dugan, C. Gariépy, J. L. Aalhus, P. Shand, and M. Juárez. 2017. Compositional and dimensional factors influencing pork belly firmness. Meat Sci. 129:54–61. doi: https://doi.org/10.1016/j.meatsci.2017.02.006

Trusell, K. A., J. K. Apple, J. W. S. Yancey, T. M. Johnson, D. L. Galloway, and R. J. Stackhouse. 2011. Compositional and instrumental firmness variations within fresh pork bellies. Meat Sci. 88:472–480. doi: https://doi.org/10.1016/j.meatsci.2011.01.029

Uttaro, B., S. Zawadski, and M. Juárez. 2020. An approach for objective and automated identification of pork belly firmness. Meat Sci. 169:108221. doi: https://doi.org/10.1016/j.meatsci.2020.108221

Uttaro, B., S. Zawadski, and M. Juárez. 2024. A new manual method for pork belly firmness measurement. MethodsX. 12:102577. doi: https://doi.org/10.1016/j.mex.2024.102577

Wei, X., B. Bohrer, B. Uttaro, and M. Juárez. 2023. Evaluating the effect of temperature and multiple bends on an automated pork belly firmness conveyor belt classification system. Meat Sci. 203:109222. doi: https://doi.org/10.1016/j.meatsci.2023.109222

Whitney, M. H., G. C. Shurson, L. J. Johnston, D. M. Wulf, and B. C. Shanks. 2006. Growth performance and carcass characteristics of grower-finisher pigs fed high-quality corn distillers dried grain with solubles originating from a modern Midwestern ethanol plant1. J. Anim. Sci. 84:3356–3363. doi: https://doi.org/10.2527/jas.2006-099