Research and analysis

Nutrient analysis of cows’ milk: sampling and analytical report

Published 6 March 2025

Summary

The Office for Health Improvement and Disparities (OHID), part of the Department of Health and Social Care (DHSC), commissioned the Food and Nutrition National Bioscience Research Infrastructure to carry out nutrient analysis of composite samples of cows’ milk bought from supermarkets in the UK. The survey was carried out in 2022 to 2023 and its aim was to provide up-to-date composition data for selected micronutrients in commonly consumed types of cows’ milk.

The last comprehensive nutrient analysis survey of cows’ milk in the UK took place in 1996. Since then, the livestock industry has made widespread changes to breeding, feeding practices, production and processing methods.

The micronutrients chosen for analysis were riboflavin, vitamin B12 and iodine. This choice was based on intelligence from the dairy industry and nutrient composition experts.

The nutrient composition data generated by this survey will be used to update the nutrient data that supports estimation of nutrient intakes in the National Diet and Nutrition Survey (NDNS). The data will be published as part of the UK food composition tables in McCance and Widdowson’s ‘The Composition of Foods’, which presents the nutrient data for the most commonly consumed foods in the UK. Additionally, the revised data will be incorporated into the next update of the composition of foods integrated dataset (CoFID) and will be added to the CoFID searchable website (email address and password required).

This project analysed 6 composite samples of pasteurised and UHT milk (ultra high temperature or ultra pasteurised milk) with different fat contents that were each made up of between 6 and 13 subsamples. As seasonal variation can affect produce where there are changes to feed throughout the year, the same types of milk were bought in both summer and winter and were analysed separately. Samples were bought from 10 different supermarkets in Norwich and prepared for analysis between August 2022 and February 2023. Details of the subsamples that make up each composite sample are provided in ‘Annex 2: nutrient analysis of cows’ milk data tables’ available on the Nutrient analysis of cows’ milk: sampling and analytical report page.

In general, lower values for riboflavin, vitamin B12 and iodine in cows’ milk were observed in the 2022 to 2023 survey compared with the 1996 survey, especially in pasteurised milk.

The differences between seasons were more consistent in the 2022 to 2023 survey, where winter samples had higher contents of all 3 nutrients compared to summer samples.

Introduction

OHID commissioned the Food and Nutrition National Bioscience Research Infrastructure to carry out a limited laboratory analysis of nutrients in cows’ milk in 2022 and 2023. The last analysis of nutrients in cows’ milk was in 1996. In consultation with dairy industry and nutrient composition experts, riboflavin, vitamin B12 and iodine were identified as the priority nutrients for analysis. This was based on intelligence about possible changes in the values of these nutrients over time.

Milk is a major source of riboflavin, vitamin B12 and iodine in the UK diet (Muehlhoff and others, 2013; Witard and others, 2022). The levels of these nutrients in milk are influenced by many factors, including:

• cow breed
• seasonal variations
• cobalt supply added to the feed (to enable vitamin B12 synthesis)
• feeding regimens

However, the precise magnitude of influence from each factor remains unclear (Matte and others, 2014).

It had been observed that UK values in CoFID for vitamin B12 and riboflavin in cows’ milk, based on the 1996 analysis, were higher in general than values recorded across non-UK food composition databases (in the USA, the Netherlands and Denmark). It was unclear whether the difference in values was due to differences in the methods used to analyse cows’ milk or whether the vitamin B12 and riboflavin content of UK cows’ milk had decreased since the 1996 analysis.

Over the past 2 decades, the dairy industry in the UK has undergone substantial changes (Barkema and others, 2015; March and others, 2014; Medeiros and others, 2022). These changes include:

• modifications in feeds
• grazing practices
• feed supplementation strategies
• genetics
• environmental conditions
• dairy processing techniques

Each of these elements has the potential to impact the riboflavin, vitamin B12 and iodine content in UK dairy products. Therefore, reanalysis of these nutrients in milk is needed to assess the contribution that milk makes to nutrient intakes in the UK population.

The nutrient composition data generated by the 2022 to 2023 survey will be used to update the nutrient data that supports estimation of nutrient intakes in the NDNS. The data will be published as part of the UK food composition tables in McCance and Widdowson’s ‘The Composition of Foods’, which presents the nutrient data for the most commonly consumed foods in the UK. Additionally, the revised data will be incorporated into the next update of the composition of foods integrated dataset (CoFID) and will be added to the CoFID searchable website (email address and password required).

Sampling and preparation

To make and prepare composite samples the following steps were taken:

1. Composite samples and the number of subsamples bought were agreed with Dairy UK and DHSC based on market shares of milk types and supermarkets.
2. Composite samples of whole, semi-skimmed and skimmed cows’ milk were analysed as these have the biggest market share across all milk types.
3. Pasteurised and UHT milks were analysed separately for each type, making a total of 6 composite samples analysed for each season (each composite sample was made up of up to 13 subsamples).
4. Subsamples were bought from 10 major supermarkets in the Norwich area according to a sample protocol produced in agreement with Dairy UK and DHSC (based on supermarket market share according to Kantar’s grocery market share data for Great Britain as of 10 July 2022). As there was no evidence for any geographical differences in the nutrient content of retail milk bought across the UK it was not considered necessary to buy samples from multiple locations.
5. Summer subsamples were bought between 10 August and 2 September 2022 and winter subsamples on 21 February 2023.
6. Subsamples were transported in cool boxes from the supermarkets and stored at 4°C before being combined in equal weights to form composite samples for analysis (preparing them as soon as possible, within the recommended use-by date).
7. Summer subsamples were prepared and processed between 11 August and 3 September 2022 and winter subsamples on 22 February 2023.
8. The subsamples were added in equal portions and combined thoroughly to create the composite sample before being split into 3 150-millilitre (ml) bottles and 2 2-litre (l) bottles.
9. Composite samples were stored at -40°C using a polystyrene box with dry ice before sending them for laboratory analysis.

Ingredients, label claims and nutrient information presented in this report have been taken directly from product packaging. Details are provided in ‘Annex 2: nutrient analysis of cows’ milk data tables’ available on the Nutrient analysis of cows’ milk: sampling and analytical report page.

Table 1: volumes of each type of cows’ milk bought from supermarkets based on their overall market share

Supermarket	Supermarket market share	Volumes of product bought
Tesco	27%	2 x 4 pints
Sainsbury’s	15%	4 pints
Asda	14%	4 pints
Morrisons	9%	2 pints and 500ml (see note)
Aldi	9%	2 pints
Lidl	8%	2 pints
Co-op	6%	2 pints
Iceland	2%	2 pints
Waitrose	5%	1 pint
Marks and Spencer (M&S)	2%	1 pint

Note: one product from Morrisons is available in a 500ml size only.

Composite sample list

The following tables outline the composite sample list, split by summer and winter samples.

Table 2: summer samples bought in 2022

Sample number	Sample name	Description
1a	Milk, whole, pasteurised	12 samples from 11 brands, chilled
1b	Milk, semi-skimmed, pasteurised	13 samples from 12 brands, chilled
1c	Milk, skimmed, pasteurised	12 samples from 11 brands, chilled
2a	Milk, whole, UHT	6 samples from 6 brands
2b	Milk, semi-skimmed, UHT	11 samples from 11 brands
2c	Milk, skimmed, UHT	10 samples from 10 brands

Table 3: winter samples bought in 2023

Sample number	Sample name	Description
1aW	Milk, whole, pasteurised	10 samples from 10 brands, chilled
1bW	Milk, semi-skimmed, pasteurised	10 samples from 10 brands, chilled
1cW	Milk, skimmed, pasteurised	11 samples from 10 brands, chilled
2aW	Milk, whole, UHT	10 samples from 7 brands
2bW	Milk, semi-skimmed, UHT	13 samples from 10 brands
2cW	Milk, skimmed, UHT	13 samples from 10 brands

Analytical methods and quality assurance

Eurofins food and feed testing laboratories were used to test the content of riboflavin, vitamin B12 and iodine in the milk. Eurofins are accredited by the UK Accreditation Service (complying with standard ISO 17025) and follow standard operating procedures to assure quality of data reported.

Riboflavin and vitamin B12 analysis was performed between August and October 2022 for summer samples and in February and March 2023 for winter samples. Iodine was analysed in March 2023 for both summer (frozen samples from summer 2022) and winter samples.

Listed below are details of the analytical methods used in the analysis of riboflavin, vitamin B12 and iodine. Vitamin B12 was analysed using 2 methods.

You can find definitions of technical terms used in this report in the glossary.

Riboflavin

The sample is dissolved in water. Internal standard is added to the sample. The sample undergoes an enzymatic treatment. Then, vitamin B1, vitamin B2, nicotinic acid, nicotinamide, vitamin B5, pyridoxine and pyridoxal are extracted using an ammonium formate solution.

The analysis is carried out using liquid chromatography with tandem mass spectrometry.

Accredited to international standard: DS EN ISO/IEC 17025 DANAK 581.

Method reference: ISO/DIN 21470:2019.

LOQ: 0.05 milligrams (mg) per 100 grams (g).

Vitamin B12 (microbiological method)

Vitamin B12 is extracted from the sample in an autoclave using a buffered solution. After dilution with basal medium (containing all required growth nutrients except cobalamin), the growth response of Lactobacillus leichmanii (ATCC 7830) to extracted cobalamin is measured turbidimetrically. This is compared to calibration solutions with known concentrations.

Accredited to international standard: DS EN ISO/IEC 17025 DANAK 581.

Method reference: Association of Official Analytical Chemists (AOAC) 952.20.

LOQ: 0.010 micrograms (μg) per 100g.

Vitamin B12 (LC-UV/DAD method)

Vitamin B12 is extracted in sodium acetate buffer in the presence of sodium cyanide (at 100°C for 30 minutes). After purification and concentration with an immunoaffinity column, vitamin B12 is determined by liquid chromatography with UV diode array detection (LC-UV/DAD) (361 nanometres (nm)).

Accredited to international standard: DS EN ISO/IEC 17025 DANAK 581.

Method reference: Campos-Gimnez and others, 2008.

LOQ: 0.080μg per 100g.

Iodine

The sample is first homogenised and digested using a mix of nitric acid and hydrogen peroxide, or an alkaline solution like potassium hydroxide to prevent iodine loss. The digested sample is then diluted and introduced into the inductively coupled plasma mass spectrometry (ICP-MS), where iodine ions are detected and quantified based on their mass-to-charge ratio.

ICP-MS accredited to international standard: BS EN ISO/IEC 17025:2017 UKAS 0342.

Method reference: internal (ICP/002).

LOQ: 0.01mg per kilogram (kg).

Results

Overall values

Values provided by Eurofins laboratories were compiled in spreadsheets for data evaluation. Where possible, analytical values were compared to previous UK food composition tables and other non-UK food composition databases (from the USA, the Netherlands and Denmark).

Riboflavin

For all pasteurised milk samples, the results for the 2022 to 2023 survey (between 0.14 to 0.21mg per 100g) were lower than the results in the 1996 survey (between 0.21 to 0.24mg per 100g) in both seasons. No seasonal data existed for UHT milks in the 1996 survey.

Comparing the averages between surveys, the new averages for all UHT milk samples (between 0.18 to 0.20mg per 100g) were slightly higher than those of the 1996 survey (between 0.17 to 0.18mg per 100g).

For all UHT milks, the results for the 2022 to 2023 survey (between 0.17 to 0.22mg per 100g) either showed no difference or were slightly higher than those of pasteurised milks (between 0.14 to 0.21mg per 100g). This observation is different from the 1996 survey, where the average riboflavin content of all pasteurised milks was higher than those of UHT milks.

Vitamin B12

Two methods were used for the analysis of vitamin B12. The results for vitamin B12 obtained through the microbiological method are listed in table 4 and described below, while the results from the LC-UV/DAD method are provided in ‘Annex 1: alternative method and results for vitamin B12 determination’.

For all pasteurised milk samples, the results for the 2022 to 2023 survey (between 0.3 to 0.5μg per 100g) were lower than the results in the 1996 survey (between 0.7 to 0.9μg per 100g) in both seasons. No seasonal data existed for UHT milks in the 1996 survey. Similar results for the previous and the new surveys were observed in UHT milks, except for skimmed milk. The UHT milk samples showed lower average levels of vitamin B12 compared to pasteurised milk in both 1996 and 2022 to 2023 analyses.

Iodine

Summer samples of all pasteurised milks showed higher iodine content (between 22 to 28μg per 100g) than in the 1996 survey (between 20 to 21μg per 100g).

However, compared to results from the 1996 survey (between 31 to 41μg per 100g), results for the 2022 to 2023 survey were lower (between 23 to 34μg per 100g), except for pasteurised skimmed milk. The same trend was found in UHT milks, where 2022 to 2023 average values were lower than those in the 1996 survey, except for skimmed milk.

No clear difference was found between the averages of UHT and pasteurised milks in the 1996 survey. In the 2022 to 2023 survey, slightly higher average iodine contents were found in UHT milks, except for semi-skimmed milk, where no difference was found.

Table 4 shows the results for riboflavin, vitamin B12 and iodine from the 2022 to 2023 and 1996 surveys for each sample type, broken down separately for summer and winter samples. Average results were calculated from summer and winter values.

Table 4: overall analytical results for riboflavin, vitamin B12 and iodine

Note: data from nutrient analysis of pasteurised liquid milk nutritional survey 1996 as published in CoFID.

Sample type	1996 riboflavin (mg per 100g) (see note)	2022 to 2023 riboflavin (mg per 100g)	1996 vitamin B12 (μg per 100g) (see note)	2022 to 2023 vitamin B12 (μg per 100g)	1996 iodine (μg per 100g) (see note)	2022 to 2023 iodine (μg per 100g)
Milk, whole, pasteurised, average	0.23	0.19	0.9	0.4	31	23
Milk, whole, pasteurised, summer	0.24	0.16	0.9	0.3	20	22
Milk, whole, pasteurised, winter	0.22	0.21	0.8	0.5	41	23
Milk, whole, UHT, average	0.18	0.19	0.2	0.2	31	27
Milk, whole, UHT, summer	Not analysed	0.18	Not analysed	0.2	Not analysed	24
Milk, whole, UHT, winter	Not analysed	0.20	Not analysed	0.2	Not analysed	30
Milk, semi-skimmed, pasteurised, average	0.24	0.18	0.9	0.4	31	26
Milk, semi-skimmed, pasteurised, summer	0.24	0.17	0.8	0.4	20	28
Milk, semi-skimmed, pasteurised, winter	0.24	0.20	0.9	0.5	41	23
Milk, semi-skimmed, UHT, average	0.18	0.20	0.2	0.3	31	26
Milk, semi-skimmed, UHT, summer	Not analysed	0.17	Not analysed	0.2	Not analysed	26
Milk, semi-skimmed, UHT, winter	Not analysed	0.22	Not analysed	0.3	Not analysed	25
Milk, skimmed, pasteurised, average	0.22	0.15	0.8	0.3	26	31
Milk, skimmed, pasteurised, summer	0.21	0.14	0.8	0.4	21	27
Milk, skimmed, pasteurised, winter	0.22	0.15	0.7	0.3	31	34
Milk, skimmed, UHT, average	0.17	0.20	0.6	0.3	25	34
Milk, skimmed, UHT, summer	Not analysed	0.18	Not analysed	0.2	Not analysed	28
Milk, skimmed, UHT, winter	Not analysed	0.22	Not analysed	0.3	Not analysed	39

Seasonal differences

In the 2022 to 2023 survey, values for all nutrients (riboflavin, vitamin B12 and iodine) were lower in pasteurised summer samples compared to winter samples, with the exception of iodine in semi-skimmed milk and vitamin B12 in skimmed milk. For UHT milk, lower values were also found in summer samples for all nutrients except iodine in semi-skimmed milk.

Interpretation

In general, lower values for riboflavin, vitamin B12 and iodine in cows’ milk have been observed in the 2022 to 2023 survey compared with the previous survey in 1996, especially in pasteurised milk.

These differences may reflect alterations in feeds, grazing practices, supplementation strategies, genetics, environmental conditions and dairy processing techniques (Barkema and others, 2015; March and others, 2014; Medeiros and others, 2022). They may also be a result of improvements in analytical methods in the 30 years since the last analysis.

The seasonal differences were more consistent in the 2022 to 2023 survey, where winter samples tend to have higher contents of all 3 nutrients compared with summer samples. This may be due to the increased supplementation of minerals in cows’ feed during the winter months, as outlined in the Agriculture and Horticulture Development Board (AHDB) report Trace element supplementation. Results for the 2022 to 2023 survey were similar to those recorded across various non-UK food composition databases.

An alternative method for measuring vitamin B12, LC-UV/DAD (Campos-Gimnez and others, 2008), was conducted in the same lab using milk samples from the same batch. The method has been validated in analyses of a range of products, including milk products. Results of the comparison between the LC-UV/DAD method and the microbiological method for each sample is presented in ‘Annex 1: alternative method and results for vitamin B12 determination’.

The LC-UV/DAD method consistently shows lower vitamin B12 contents in most milk types compared to the microbiological method, both in winter and summer. Similarly, as noted with the microbiological method, UHT milk had lower levels of vitamin B12 compared to pasteurised milk varieties. Furthermore, fewer differences between winter and summer samples were found in pasteurised milk when using the LC-UV/DAD method compared to the microbiological method.

In the UK, estimates of vitamin B12 intakes in the NDNS have been based on analysis of foods using the microbiological method (as presented in table 4). However, newer chromatographic methods (as presented in ‘Annex 1: alternative method and results for vitamin B12 determination’) are faster and more reliable for analysing vitamin B12 content in food and are more relevant for human use (Campos-Gimnez and others, 2008; Chamlagain and others, 2015).

It is likely that future vitamin B12 analysis in foods to support estimation of intakes will move towards chromatographic methods.

Glossary

Accredited to

Refers to the official recognition or certification granted to a laboratory or institution, confirming that it meets the required standards for competence in a specific field of testing or calibration.

Autoclave

A device that sterilizes equipment and materials using high-pressure steam at around 121°C, commonly used to kill microorganisms.

Basal medium

A simple nutrient medium that supports basic microbial or cell growth without added supplements.

BS EN ISO/IEC 17025:2017 UKAS 0342

BS: adoption of ISO/IEC 17025 by the UK.

EN: European Standard.

ISO/IEC 17025: the international standard for laboratory testing and calibration, ensuring that laboratories are well-equipped to produce accurate and trustworthy results.

2017: refers to the latest version of the standard.

UKAS 3042: the accreditation number issued by UKAS (UK Accreditation Service), the official accreditation body in the UK.

Buffered solution

A solution that maintains a stable pH when small amounts of acid or base are added, usually containing a weak acid and its conjugate base.

Cobalamin

Also known as vitamin B12, essential for DNA synthesis, nerve function and red blood cell formation.

DS EN ISO/IEC 17025 DANAK 581

DS: adoption of ISO/IEC 17025 by Denmark.

EN: European Standard.

ISO/IEC 17025: the international standard for laboratory testing and calibration, ensuring that laboratories are well-equipped to produce accurate and trustworthy results.

DANAK 581: the accreditation number issued by DANAK (the Danish Accreditation Fund), the official accreditation body in Denmark.

Homogenised

A process where different components are mixed or blended to create a uniform and consistent mixture or composition.

Immunoaffinity column

A chromatography tool that uses antibodies to selectively capture and isolate specific proteins or molecules from a mixture.

Inductively coupled plasma mass spectrometry (ICP-MS)

An analytical technique used to detect and measure trace elements and isotopes in a sample by ionising it with an inductively coupled plasma and analysing the ions using mass spectrometry.

Lactobacillus leichmanii (ATCC 7830)

A bacterial strain used in bioassays, particularly for measuring cobalamin (vitamin B12) levels.

Bioassays are used to test or analyse the biological activity of a substance by studying the substance’s effect on an organism or in a test tube under controlled conditions.

Liquid chromatography

A technique to separate, identify and quantify compounds in a liquid sample by passing it through a packed column under pressure.

LOQ (limit of quantification)

The lowest concentration of a substance that can be reliably quantified with accuracy and precision using a specific analytical method.

Mass spectrometry and tandem mass spectrometry

An analytical technique used to identify and quantify compounds based on the mass-to-charge ratio of their ions. Tandem mass spectrometry involves multiple stages of mass analysis to provide more detailed structural information.

Mass-to-charge ratio

A measure used in mass spectrometry to describe the ratio of the mass of an ion to its electrical charge.

Turbidimetrically

A method for measuring particle concentration by assessing the cloudiness of a solution based on light scattering.

UV diode array detection

A detection method in liquid chromatography that measures the absorbance of UV light across multiple wavelengths to analyse compounds.

Acknowledgements

This work was carried out by Food and Nutrition National Bioscience Research Infrastructure at the Quadram Institute Bioscience, funded by the Department of Health and Social Care and the Biotechnology and Biological Sciences Research Council (BBSRC) core capability grant (number BB/CCG2260/01).

The authors gratefully acknowledge the helpful discussions with Dairy UK and the support from BBSRC.

References

Barkema HW, von Keyserlingk MA, Kastelic JP, Lam TJ, Luby C, Roy JP, LeBlanc SJ, Keefe GP and Kelton DF. Invited review: changes in the dairy industry affecting dairy cattle health and welfare. Journal of Dairy Science 2015: volume 98, issue 11, pages 7,426 to 7,445.

Campos-Gimnez E, Fontannaz P, Trisconi MJ, Kilinc T, Gimenez C and Andrieux P. Determination of vitamin B12 in food products by liquid chromatography/UV detection with immunoaffinity extraction: single-laboratory validation. Journal of AOAC International 2008: volume 91, issue 4, pages 786 to 793.

Chamlagain B, Edelmann M, Kariluoto S, Ollilainen V and Piironen V. Ultra-high performance liquid chromatographic and mass spectrometric analysis of active vitamin B12 in cells of Propionibacterium and fermented cereal matrices. Food Chemistry 2015: volume 166, pages 630 to 638.

Matte JJ, Britten M and Girard CL. The importance of milk as a source of vitamin B12 for human nutrition. Animal Frontiers 2014: volume 4, issue 2, pages 32 to 37.

March MD, Haskell MJ, Chagunda MGG, Langford FM and Roberts DJ. Current trends in British dairy management regimens. Journal of Dairy Science 2014: volume 97, issue 12, pages 7,985 to 7,994.

Medeiros I, Fernandez-Novo A, Astiz S and Simões J. Historical evolution of cattle management and herd health of dairy farms in OECD countries. Veterinary Sciences 2022: volume 9, issue 3, page 125.

Muehlhoff E, Bennett A and McMahon D. Milk and dairy products in human nutrition (PDF, 3MB). Food and Agriculture Organization of the United Nations, 2013.

Witard OC, Bath SC, Dineva M, Sellem L, Mulet-Cabero AI, van Dongen LH, Ju-Sheng Z, Valenzuela C and Smeuninx B. Dairy as a source of iodine and protein in the UK: implications for human health across the life course, and future policy and research. Frontiers in Nutrition 2022: volume 9, article number 800559.

Annex 1: alternative method and results for vitamin B12 determination

Vitamin B12 is extracted in sodium acetate buffer in the presence of sodium cyanide (100°C, for 30 minutes). After purification and concentration with an immunoaffinity column, vitamin B12 is determined by LC-UV/DAD (361 nm).

Accredited to: DS EN ISO/IEC 17025 DANAK 581.

Method reference: Campos-Gimnez and others, 2008.

LOQ: 0.080μg per 100g.

Figure 1: comparisons of vitamin B12 values (μg per 100g) in all milk composition samples between LC-UV/DAD and microbiological methods

Season and method	Milk, whole, pasteurised	Milk, semi-skimmed, pasteurised	Milk, skimmed, pasteurised	Milk, whole, UHT	Milk, semi-skimmed, UHT	Milk, skimmed, UHT
Summer LC-UV/DAD	0.28	0.28	0.29	0.14	0.14	0.14
Summer microbiological	0.35	0.35	0.36	0.17	0.19	0.17
Winter LC-UV/DAD	0.31	0.31	0.25	0.18	0.19	0.18
Winter microbiological	0.52	0.47	0.31	0.21	0.25	0.34
Average LC-UV/DAD	0.29	0.3	0.27	0.16	0.17	0.16
Average microbiological	0.44	0.41	0.34	0.19	0.22	0.25