A reassessment of pork nitrogen factors

Abstract

The Analytical Methods committee has received and approved the following report from the nitrogen factors sub-committee.

---

Summary

Labelling rules require that where ingredients are highlighted in the name of the food or are associated with that food, the amount of that ingredient must be declared as a percentage of the final product (QUID – quantitative ingredient declaration). The rules also require that where a product looks like a cut or joint of meat or poultry e.g. a cured pork leg with added water, if there is more than 5% added water, this has to appear in the name of the food. Nitrogen factors are analytical factors used to measure meat and fish content (i.e. the amount of meat or fish used) in compound products, as well as added water in raw meat and fish and their products (e.g. cooked or cured). They are used by both enforcement laboratories (public analysts) to check (QUID) labelling claims on meat and fish content, and by industry to check their raw materials and finished products.

The UK market value of pork and pork products is around £4.4 billion. In addition to home production, a substantial amount of pork and cured pork (ham or bacon) is imported to the UK from Europe especially Denmark and the Netherlands.

The last extensive AMC study determining pork nitrogen factors of different joints and the overall carcase from UK pigs only, was published in 1991. Over the past 21 years, the genetics, breeds and methods of rearing pigs have changed. In addition, the average UK carcase weights have also increased. In the 1991 study, pigs used in the study were killed at 50–70 kg. Today, the carcase weight for UK pigs is usually higher (78–80 kg). Danish pigs are slaughtered around 82 kg and Dutch pigs at 85 kg. In this study, the nitrogen factors for the whole side, four main joints of pork (belly, loin, shoulder and leg), and individual shoulder (hand and collar) and leg joints (silverside, topside, shin, chump and flank) from pigs reared in the UK, Denmark and the Netherlands were reassessed, and compared to the values of nitrogen factors, where relevant, in the 1991 study.

A standard operating procedure (SOP) was agreed on preparing deboned and rind removed joints and cuts from a whole side. Each country (UK, DK, NL) supplied a total of 40 commercial sides in 2 sampling sessions – March 2013, and October/December 2013. 705 samples were prepared according to the SOP, and analysed for fat, ash, moisture and nitrogen in 3 accredited labs using BSI/ISO methodology or equivalent.

The average weight of carcases (84 kg) used in this study was much heavier than the average national UK carcase weight in the 1991 study (70 kg). The UK and Danish median carcase weight was 81 kg, whereas the median Dutch carcase weight was 89 kg. The overall % lean of the carcases was 61%, which is the highest category in the EU carcase classification scheme. Gender had no effect on the proximate content of the samples. There was a variation in fat free nitrogen (N_FF) along the loin between 3.6 and 3.8, when sampled from the shoulder to the leg end. Unlike the 1991 study, there was no strong correlation between N_FF, carcase weight and % lean (or its indicator P₂ fat thickness).

There were country differences for certain joints or cuts, but these were not consistent over all the joints for each country. When the N_FF values were combined for whole side N_FF in proportion to the joint/cuts in the side, the differences between the countries were quite small. The recommended whole side factor in this study is 3.60, which is higher than the 3.50 recommended in the 1991 study.

Report

The membership of the nitrogen factors sub-committee responsible for the preparation of this report was: Dr Mark Woolfe, (Chairman), Dr Roger Wood (Secretary), Mr Andrew Caines, Ms Bhavna Parmar, Ms Pendi Najran, Mr Charles. Boardman, Mr David Keeble, Mr Duncan Campbell, Mr Jeremy Hall, Prof Michael Thompson, Mr Robbie Beattie, Ms Selvarani. Elahi, Mr Steve Lamming, Mr Steve Moore, Ms Henrietta Sameke, Mr Andrew Furmage, Ms Liz Moran.

Participation in the study was by Dr Mark Woolfe, Mr Andrew Caines, Ms Jo Hogg, Ms Jackie Carter, Mr David Selby, Mr Nick Goodwin, Ms Gayna Quinn, Ms Margrethe Jensen, Dr Ronald Klont, Mr Sjaak Wisse, Ms Selvarani Elahi, Ms Kirstin Gray and Mr Stephen Ellison, who was responsible for the statistical evaluation.

Introduction

The food information regulation¹ requires that where ingredients are highlighted in the name of the food or are associated with that food, the amount of that ingredient must be declared as a percentage of the final product (QUID – quantitative ingredient declaration). Although the quantity of an ingredient is calculated on a recipe basis, enforcement authorities usually check the declaration by analysis of the finished product. The analysis determines the nitrogen (mainly on a fat free basis N_FF) of meat ingredient, and converts this to a meat content using a previously determined nitrogen factor.

Where relevant, the determination of meat content permits the calculation of added water by difference. The same regulation¹ requires that where a product looks like a cut or joint of meat or poultry e.g. a cured pork leg (ham) with added water, if there is more than 5% added water, then “added water” has to appear in the name of the food. Also if ingredients are used from a different species e.g. hydrolysed proteins or gelatine as ingredients, then they also have to be declared in the name of the food. It also requires country of origin to be given for all meat and poultry cuts. The European Commission has yet to decide whether new rule will make existing government guidance mandatory, which recommends that in the case of bacon or ham prepared in the UK, the origin of the pork is stated e.g. British bacon prepared from Dutch pork.

In 2014, the provisional total UK market sales for pork and pork products were just over £4.4 billion.² UK production of pork and pork products was provisionally given as £3.2 billion in 2013.³ In 2014, the UK imported £699 million worth of pork, nearly all (99.6%) from the EU, of which 23% came from Denmark, 23% from Germany, and 13% from The Netherlands. In the same year, the UK imported £601 million worth of bacon and ham, of which 39% came from Denmark, 34% came from the Netherlands, and 13% came from Germany.⁴ The figures show the importance of pork and pork products on the UK market, as well as the importance of pork, bacon and ham from Denmark, and the Netherlands.

The last extensive study on pork nitrogen factors was funded by MAFF and carried out by the RSC/AMC-NFSC in cooperation with the pork industry in 1991.⁵ This gave recommended nitrogen factors for five joints (belly, loin, collar, hand, and leg) of pork and the whole side. It is also gave correlation curves for different carcase weights with P₂ fat thickness (EC carcase classification scheme). However, since 1991 the genetics of pigs have changed, and the main commercial genotypes are Landrace crosses. Leaner and faster growing pigs are now raised and slaughtered around 78–80 kg weight compared to their Danish and Dutch counterparts, which are slaughtered at 82 and 90 kg weight respectively. There is also more outdoor rearing of pigs.

Given the market importance of pork and pork products, a reassessment of pork nitrogen factors was overdue. In addition given the supply importance of pork from the Netherlands and Denmark, a comparison of UK, Dutch and Danish pork was requested by the UK industry in order to give more accurate meat contents especially in the light of country of origin requirements.

Experimental

Pork samples

A sampling SOP (standard operating procedure) was agreed between the project participants on sampling the pork sides. The pork sides were to be cut into the 4 main joints – loin, belly, shoulder and leg (see Fig. 1), the rind removed from the joints leaving the subcutaneous fat, and each joint deboned and dissected into sections or muscle cuts in order to take representative samples. It was agreed at the industry's request that the shoulder would be prepared as in the 1991 study i.e. collar and hand, but the leg joint would be cut into 5 muscle cuts – topside, silverside, chump, flank and shin. Each company preparing the samples provided deboned weight data of the proportion of hand and collar in the shoulder joint, the proportion of silverside, topside, shin, chump and flank in the leg joint, and the proportion of loin, belly, shoulder and leg joint in the whole side.

Fig. 1 Standardised joint classification of a pork carcase.

Samples were prepared from a total of 40 pork sides from each country – UK, Denmark and the Netherlands. These were divided into two equal sampling sessions – March 2013 and Oct/Dec 2013. The samples were prepared in Denmark on the Danish samples, and in two English plants, which prepared the UK samples. One company produced both UK and the Dutch samples. The samples were distributed to the 3 accredited laboratories participating in the project. Sides were taken from commercial runs in the abattoirs, and from the data provided, were a distribution of genders, genotypes and weights reflecting the market in each country. In the UK, modern genotype breeds are Large White/Landrace crosses with other breeds such as Hampshire, Pietrain and Duroc, whereas Denmark and the Netherlands use national Landrace breeds. Table 1 gives the overall number of samples prepared and analysed. 705 results were used for samples following the instructions in the SOP. Table 2 shows the proportion (% by weight) of all the joints and cuts in the deboned, rindless whole side. The mean weights for all the cuts, joints and whole sides from the UK, Denmark and the Netherlands were provided for all the sides used in the study. The two UK companies gave separate weight data for their half number of sides, and the reported proportions were averaged across the two UK suppliers to provide the averaged UK proportions. There are some differences between countries, but these may be more due to the way that sides have been dissected rather than significant differences in the conformation of the different breeds.

Table 1 Actual number of pork samples prepared and analysed (no. of planned samples)

Joint or cut	Denmark	The Netherlands	United Kingdom
Belly	40 (40)	40 (40)	45 (40)
Loin	40 (40)	40 (40)	45 (40)
Shoulder – hand	20 (20)	20 (20)	23 (20)
Shoulder – collar	20 (20)	20 (20)	22 (20)
Leg – silverside	40 (40)	39 (40)	30 (40)
Leg topside	20 (20)	20 (20)	15 (20)
Leg – chump	20 (20)	20 (20)	15 (20)
Leg – flank	20 (20)	21 (20)	15 (20)
Leg – shin	20 (20)	20 (20)	15 (20)
Total	240 (240)	240 (240)	225 (240)

---

Table 2 Percentage by weight of the whole side for all joints and cuts for the 3 countries

Joint/cut	Country
	DK	NL	UK^a
a Average of the two UK companies.
Belly	16.7	18.0	16.7
Loin	21.2	19.0	18.8
Shoulder – collar	13.9	16.3	13.1
Shoulder – hand	15.9	18.1	19.1
Leg – chump	4.4	3.3	2.6
Leg – flank	5.7	5.5	6.7
Leg – shin	2.7	3.0	3.0
Leg – silverside	11.1	8.7	10.5
Leg – topside	8.5	8.2	9.5

---

Methods of analysis

Samples of each joint or muscle cut were analysed in duplicate using the following chemical analysis methods based on British standards institution, Analytical Methods for meat and meat products.

Part 1 1970 (1993) determination of ash⁶

Part 2 1980 (1993) determination of nitrogen or equivalent method.⁷ All labs used an equivalent method by determining nitrogen by Dumas combustion using a LECO 2000 CNS.

Part 3 1970 (1997) determination of moisture⁸

Part 4 1970 (1993) determination of total fat⁹

The detailed requirements for analysis included:

Two replicates of the four analyses were required per sample, which should add up to 100% ± 2%. The duplicates were randomised within each batch. Analytical agreement was required between the duplicates – the repeatability limits were 0.5 g/100 g (moisture and fat) and 0.1 g/100 g (nitrogen and ash).

The laboratories analysed a standard reference material (e.g. FAPAS meat sample or LGC Reference Material 7152, Processed Pork) in duplicate with each batch of samples. Laboratories assessed their performance against the above limits, and decided whether repeats were necessary. All three laboratories used the rapid Dumas method for nitrogen determination. This measures the non-protein nitrogen as well as the protein nitrogen, and hence gives higher results than the Kjeldahl method.⁶ The difference is small, and Dumas is higher by a factor of 1.014,¹⁰ which is more important for fish than meat, and was therefore not taken into account in this study.

The results were analysed statistically to see whether there were any significant differences for gender, weight of carcase, lean meat percentage, country, and joint or cut. The results were also compared to those in the 1991 study where relevant.

Results

Carcase weight and percentage lean meat of carcases

In the 1991 study, the P₂ fat thickness was used as the indicator for the EU pig carcase classification. The EU classification is based on percentage lean of the dressed carcase after slaughter, and the P₂ fat thickness is in fact just one of the methods used to calculate the % lean.¹¹ The pig data from Denmark and the Netherlands provided the % lean of the carcase, whereas in the UK the P₂ fat thickness was given. One UK company gave both, and hence it was possible to use this information as a data set to convert the P₂ fat thickness and carcase weight to % lean (Table 3), and this correlation was used to convert the P₂ fat thickness to % lean (see Fig. 2).

Table 3 Relationship between % lean and P₂ fat thickness and carcase weight

Term	Coefficient	Std. error
UK data only.
Intercept	66.4439	0.2528
P2 (mm)	−0.946	0.0125
Weight (kg)	0.0675	0.0029

---

Fig. 2 Predicted % lean versus observed % lean (UK data).

The carcase weight distribution of the UK, Danish and Dutch pigs used in this study is shown in Fig. 3. The weights were taken as representative of the market in these countries. In the 1991 study, the results were adjusted to the national average carcase weight of 70 kg. In this study, the UK carcase weights had the widest range (60–104 kg) with the median of 81 kg, which was the same median weight (81 kg) as the Danish pigs with a much narrower range (70–88 kg). The Dutch pigs were heavier at a median weight of 89 kg (range 83–104 kg). The overall median weight of the pigs used in this study was 84 kg.

Fig. 3 Pig carcase dressed weight distribution in the UK, the Netherlands and Denmark.

Fig. 4 shows the distribution of % lean of the sides used in this study. The UK sides had a wider distribution than the sides from Denmark, but had a leaner median of 61.5% lean similar to the Danish sides. Whereas the heavier Dutch pig sides had a lower median % lean of 59.6. The overall median % lean of the sides used in this study was 61.

Fig. 4 % Lean distribution of carcases from the UK, the Netherlands and Denmark.

Analytical results of all pork samples

Tables 4–7 give the mean proximate content in g/100 g for ash, fat, moisture and nitrogen by joint/cut and country. Table 8 gives the mean fat free nitrogen (N_FF) by cut and country, and reduces the variability due to different fat levels.

Table 4 Mean ash content by cut and country

Joint	Cut	Country
		DK	NL	UK
In Tables 4–8, all values are in g/100 g. Standard errors are given in parentheses. Means sharing a superscript (a and/or b) in each row are NOT significantly different at the 95% level of confidence (pairwise test with correction for multiple comparisons).
Belly	—	0.82 (0.02)	0.77 (0.02)	0.88 (0.02)
Loin	—	0.88^a (0.03)	0.92^ab (0.03)	0.93^b (0.03)
Shoulder	Collar	0.91^a (0.03)	0.95^a (0.03)	1.02 (0.03)
Shoulder	Hand	0.98^a (0.02)	0.96^a (0.02)	0.96^a (0.02)
Leg	Chump	1.13^ab (0.01)	1.10^a (0.01)	1.14^b (0.01)
Leg	Flank	1.12^a (0.02)	1.07^b (0.02)	1.12^ab (0.02)
Leg	Shin	1.06^a (0.02)	1.05^a (0.02)	1.04^a (0.02)
Leg	Silverside	0.98 (0.02)	1.04^a (0.02)	1.05^a (0.02)
Leg	Topside	1.13^a (0.02)	1.12^a (0.02)	1.12^a (0.02)

---

Table 5 Mean fat content by cut and country

Joint	Cut	Country
		DK	NL	UK
See footnote for Table 4.
Belly	—	25.36^a (1.53)	28.48^a (1.53)	20.91 (1.50)
Loin	—	22.55 (1.08)	18.11^a (1.08)	18.69^a (1.02)
Shoulder	Collar	18.25^a (3.14)	15.98^a (3.14)	11.77 (3.12)
Shoulder	Hand	14.17^a (1.04)	13.73^a (1.04)	15.95^a (0.98)
Leg	Chump	2.19^a (0.19)	2.93^b (0.19)	2.54^ab (0.22)
Leg	Flank	2.43^a (0.35)	2.97^a (0.35)	3.06^a (0.37)
Leg	Shin	3.09^a (0.64)	3.93^ab (0.64)	4.49^b (0.64)
Leg	Silverside	12.13 (1.20)	5.37 (1.20)	7.72 (1.21)
Leg	Topside	2.60^a (0.33)	2.27^a (0.33)	2.67^a (0.34)

---

Table 6 Mean moisture content by cut and country

Joint	Cut	Country
		DK	NL	UK
See footnote for Table 4.
Belly	—	57.59 (1.25)	54.89 (1.25)	60.70 (1.23)
Loin	—	59.03 (0.77)	62.13^a (0.77)	61.67^a (0.73)
Shoulder	Collar	63.74^a (1.98)	65.20^a (1.98)	68.47 (1.97)
Shoulder	Hand	66.45^ab (0.85)	67.36^a (0.85)	64.47^b (0.79)
Leg	Chump	74.68^a (0.51)	73.31^b (0.51)	73.93^ab (0.53)
Leg	Flank	75.70^a (0.52)	75.11^a (0.52)	74.85^a (0.53)
Leg	Shin	74.54^a (0.81)	74.39^a (0.81)	73.20 (0.81)
Leg	Silverside	67.40 (0.80)	72.33 (0.80)	70.25 (0.81)
Leg	Topside	74.61^a (0.53)	74.60^a (0.53)	73.43 (0.54)

---

Table 7 Mean nitrogen content by cut and country

Joint	Cut	Country
		DK	NL	UK
See footnote for Table 4.
Belly	—	2.63^a (0.04)	2.57^a (0.04)	2.88 (0.04)
Loin	—	2.83 (0.08)	3.10^a (0.08)	3.00^a (0.08)
Shoulder	Collar	2.76 (0.10)	2.95^a (0.10)	3.07^a (0.10)
Shoulder	Hand	3.00^a (0.05)	2.93^a (0.054)	2.97^a (0.05)
Leg	Chump	3.57^a (0.058)	3.65^a (0.058)	3.60^a (0.06)
Leg	Flank	3.39^a (0.04)	3.38^a (0.04)	3.416^a (0.04)
Leg	Shin	3.41^ab (0.05)	3.30^a (0.05)	3.45^b (0.05)
Leg	Silverside	3.12 (0.08)	3.44^a (0.08)	3.36^a (0.08)
Leg	Topside	3.51^a (0.04)	3.59^ab (0.04)	3.62^b (0.04)

---

Table 8 Mean fat free nitrogen content by cut and country

Joint	Cut	Country
		DK	NL	UK
See footnote for Table 4.
Belly	—	3.52 (0.05)	3.60^a (0.05)	3.60^a (0.05)
Loin	—	3.65^a (0.06)	3.78 (0.06)	3.68^a (0.06)
Shoulder	Collar	3.40 (0.04)	3.53^a (0.04)	3.48^a (0.04)
Shoulder	Hand	3.50^a (0.05)	3.40 (0.05)	3.53^a (0.05)
Leg	Chump	3.64^a (0.07)	3.76^a (0.07)	3.70^a (0.07)
Leg	Flank	3.47^a (0.05)	3.49^a (0.05)	3.52^a (0.05)
Leg	Shin	3.51^ab (0.07)	3.43^a (0.07)	3.62^b (0.07)
Leg	Silverside	3.56^a (0.05)	3.64^ab (0.05)	3.65^b (0.05)
Leg	Topside	3.60^a (0.05)	3.68^ab (0.05)	3.72^b (0.05)

---

Effect of carcase weight and percentage lean meat on the fat free nitrogen content (N_FF)

In the 1991 study, there were significant carcase weight effects on N_FF values with a quadratic dependence on P₂ fat thickness (an indicator of lean meat content). In this study, there were no consistent trends for N_FF with carcase weight within cuts. A weak quadratic relationship across all cuts proved to be statistically significant at the 95% level of confidence and is shown in Fig. 5; this can, however, be attributed to a small number of lower nitrogen values at extreme carcase weights. Unlike the 1991 study, the present study found no statistically significant quadratic dependence of nitrogen content on percentage lean meat or with P2 thickness inferred from % lean using the relationship in Table 3.

Fig. 5 Fat free nitrogen content by carcase weight, grouped by cut points are mean N_FF content for each sample tested. The dashed line shows the quadratic general dependence found, adjusted for differences between cuts.

Effect of gender on proximates

Information about the gender of the pigs was provided for all the samples. Four descriptions were reported – boar, castrate (male), female and gilt (non-breeding female), leaving the breeding status for some samples unclear. It was assumed that breeding boars and sows were unlikely to form a large fraction of stock sold for meat, and the analysis was done using just male and female categories. There is no significant effect of gender on fat, moisture and fat free nitrogen, and no statistically significant effect of gender was found.

Laboratory effect on proximate content measurements

Analysis of the laboratory variation between the 3 labs for moisture and fat free nitrogen for the results of samples of loin, belly, collar and hand show a significant laboratory effect. Compared to the variation between individual samples, the effect of the laboratory on the results was not practically significant. However, the statistical modelling and the resulting standard errors shown in Tables 4–8 allowed for laboratory effects as a random effect on results, where more than one laboratory reported results for a given cut.

Effect of location on the loin on its nitrogen and fat free nitrogen content

The sampling of the loin in the SOP was designed specifically to show whether there was any differences in proximate composition with location along the loin from the shoulder end to the leg end. The loin is the largest joint in the side, and it was divided into 8 approximately equal parts. The only proximate where a relationship was found was with the nitrogen and the N_FF content, and nitrogen content was modelled as a function of location. Table 9 and Fig. 6 show that it was possible to fit a quadratic dependence of nitrogen and N_FF to loin location. The fitted model in Fig. 6 shows that the N_FF starts at 3.6 g/100 g at the shoulder end and rises to a maximum of 3.8 around the 5^th and 6^th location along the loin, and dropping down to 3.7 at the leg end.

Table 9 Loin nitrogen and fat free nitrogen content on location

Coefficient	Nitrogen (standard error)	Nitrogen NFF (standard error)
Intercept b0	2.83 (0.06)	3.60 (0.04)
Location b1	0.72 (0.29)	0.60 (0.19)
Location b2	−0.56 (0.29)	−0.50 (0.19)
Regression R^2	0.066	0.094

---

Fig. 6 Nitrogen and fat-free nitrogen (NFF) content along loin joint.

Table 9 shows coefficients for the model N = b₀ + b₁L + b₂L², where N is the percentage nitrogen and L is the proportion of total loin length from shoulder to leg (i.e. L takes values in the range [0–1] corresponding to reported SOP location values in the range 1–8).

In Fig. 6 above, points show all mean N and N_FF% content for loin samples. The horizontal axis shows loin location as a fraction of the distance from shoulder to leg, based on the locations outlined in the SOP; i.e. the range 0.0–1.0 corresponds to the SOP range 1–8. The solid line is a smoothing line through the points, formed from a locally weighted regression fitted to the data. The dashed red line is the best fit second order polynomial.

Calculation of N_FF of the whole side, leg and shoulder from the individual cuts

Using Table 2, the mean N_FF values (g/100 g) were calculated for all the main joints and whole side for the 3 countries. The standard errors in Table 10 were calculated from those in Table 8 using a simple square root of the weighted sum of variances. Note that this does not allow for correlation between results for different cuts, and the standard errors are accordingly likely to be somewhat underestimated. Table 10 shows the N_FF values for the main joints and whole side, and compares them to the N_FF values for the lean, inter-muscular and subcutaneous fat values of the 1991 study. Again there are differences in N_FF values between the 3 countries but these are relatively small and are not consistent for one particular country. Comparison with 1991 study shows that all the joints and cuts have an increased value for N_FF compared to the 1991 study, but the differences are relatively small between 0.05 and 0.1.

Table 10 Mean fat free nitrogen content (standard error in brackets) of main joints and whole side for UK, NL and DK

Joint	Country
	DK	NL	UK	UK 1991 study
Belly	3.52 (0.05)	3.60 (0.05)	3.60 (0.05)	3.51
Loin	3.65 (0.06)	3.78 (0.06)	3.68 (0.06)	3.66
Collar	3.40 (0.04)	3.53 (0.04)	3.48 (0.04)	3.42
Hand	3.50 (0.05)	3.40 (0.05)	3.53 (0.05)	3.44
Shoulder	3.45 (0.03)	3.46 (0.03)	3.51 (0.03)	—
Leg	3.56 (0.03)	3.61 (0.03)	3.65 (0.03)	3.50
Whole side	3.54 (0.02)	3.59 (0.02)	3.60 (0.02)	3.52

---

Conclusions and recommended nitrogen factors

The average weight of carcases (84 kg) used was much heavier than the average national UK carcase weight in the 1991 study (70 kg). The UK and Danish median carcase weight was 81 kg, whereas the median Dutch carcase weight was 89 kg. The overall % lean of the carcases was 61%, which is the highest category in the EU classification scheme. The heavier pigs used at the moment also have lower fat content. In the 1991 study, the whole side from the UK average pig of 70 kg had a fat content of lean, inter-muscular and subcutaneous fat of 20.9 g/100 g, whereas the fat content of the whole side from the median UK pig of 81 kg in this study is only 13.1 g/100 g (calculated from the proportion data in Table 2).

Gender had no effect on the proximate content of the samples. There was a small variation in fat free nitrogen along the loin when sampled from the shoulder to the leg end. The N_FF varied between 3.6 and 3.8 with the maximum around halfway to three quarters of the way along the loin. Unlike the 1991 study, there was no strong correlation between N_FF, carcase weight and % lean (or its indicator P₂ fat thickness).

There were country differences for certain joints or cuts, but these were not consistent for any one country. When the N_FF values were combined for whole side N_FF in proportion to the joint/cuts in the side, the differences between the countries were quite small. This would indicate that a major part of the variation was probably due to differences in removing the rind, deboning and dissecting the whole side by the companies preparing the samples.

Table 11 gives the recommended nitrogen factors for all the joints, cuts and whole side. The recommended factors for the different countries have been obtained simply by rounding up or down the values to the nearest 0.05 from Tables 8 and 10. As in most cases the differences are small, an average factor from all three countries is also given.

Table 11 Recommended nitrogen factors for joint/cut and country

Joint	Cut	Country
		DK	NL	UK	All 3 countries (standard error)	1991 study for 70 kg carcase
Values are g/100 g and rounded to nearest 0.05.
Belly	—	3.50	3.60	3.60	3.55 (0.05)	3.50
Loin	—	3.65	3.75	3.70	3.70 (0.06)	3.65
Shoulder	—	3.45	3.45	3.50	3.45 (0.03)	—
Shoulder	Collar	3.40	3.55	3.50	3.45 (0.04)	3.40
Shoulder	Hand	3.50	3.40	3.55	3.50 (0.05)	3.45
Leg	—	3.55	3.60	3.65	3.60 (0.03)	3.50
Leg	Chump	3.65	3.75	3.70	3.70 (0.07)	—
Leg	Flank	3.45	3.50	3.50	3.50 (0.05)	—
Leg	Shin	3.50	3.45	3.60	3.50 (0.07)	—
Leg	Silverside	3.55	3.65	3.65	3.60 (0.05)	—
Leg	Topside	3.60	3.70	3.70	3.65 (0.05)	—
Whole side	—	3.55	3.60	3.60	3.60 (0.02)	3.50

---

In Table 11, the recommended N_FF values for all 3 countries are compared to recommended factors in the 1991 study. It can be seen that the values only differ by 0.05 or a maximum of 0.10. In particular, the recommended factor for the whole side has change from 3.50 to 3.60, which represents around 2–3% lean meat protein for a 80% or more pork ingredient in a pork product.

Acknowledgements

This study was undertaken as part of the programme of work of the Royal Society of Chemistry, Analytical Methods Committee – Nitrogen Factors Sub-Committee. The project was funded by Defra as part of its Food Authenticity Programme. Thanks are due to the companies who supplied the pork sides and dissected them into samples, for doing so in their very busy production schedules.

References

1. Regulation (EU) No 1169/2011 of the European Parliament and of the Council of 25 October 2011 on the provision of food information to consumers, OJ L304, 22 November 2011, pp. 18–63.
2. Kantar Worldpanel – GB consumers expenditure on pork, pork sausages, and pork/ham sliced cooked meats for the 52 weeks ending 9 November 2014. Taken from BPEX (British Pig Executive) Meat Market Summary –http://www.bpex.org.uk/prices-stats/consumption/meat-market-summary/.
3. PRODCOM provisional estimates for 2013 – Division 10: Meat Values by Customs Codes – Office for National Statistics (ONS).
4. Pork and Ham Imports for 2014 – Defra Trade Statistics Unit; HM Revenue & Customs.
5. Pork Nitrogen Factors – A Reassessment, Analyst, July 1991, vol. 116, pp. 761–6.
6. BS4401-1: 1998, ISO 936:1998. Methods of Test for Meat and Meat Products. Part 1: Determination of Total Ash.
7. British Standard, BS 4401:1980 ISO 937:1978, “Methods of Test for Meat and meat products – Part 2: Determination of nitrogen content (reference method)”.
8. BS4401-3: 1997, ISO 1442:1997. Methods of Test for Meat and Meat Products. Part 3: Determination of Moisture Content (Reference Method).
9. BS 4401: Part 4: Method A 1970 (Weibull Stoldt, acid hydrolysis) UKAS accredited method based on BS4401-4: 1970. Methods of Test for Meat and Meat Products. Part 4: Determination of Total Fat Content: Method A (Weibull Stoldt).
10. M. Thompson, L. Owen, K. Wilkinson, R. Wood and A. Damant, A Comparison of the Kjeldahl and Dumas Methods for the Determination of Protein in Foods, using Data from a Proficiency Testing Scheme, Analyst, 2002, 127, 1666–1668.
11. Pig Carcase Classification for the Meat Industry – Meat and Livestock Commercial Service Ltd’s (MLCSL) Independent Authentication Service. http://www.mlcsl.co.uk/pdf/PigCarcaseClassificationFINAL23Aug.pdf.

---

Footnote

† Correspondence should be addressed to Dr M. Woolfe, e-mail: mjwoolfe@gmail.com.

---