Milk of the future?

New selection criteria in our breeding programs could make butterfat healthier for humans, enhancing returns

The possibility of improving milk composition by genetic selection is appealing, and may have big implications for human health. It is quite possible to select for higher levels of desirable fatty acids (FAs) in the mix that makes up milk's butterfat component, a recently published research paper from the Netherlands suggests.

This points the way to possibly changing butterfat composition to produce a product with greater health benefits. For example, you might breed for cows to produce a higher level of unsaturated fatty acids or greater concentration of conjugated linoleic acid (CLA).

Butterfat already contains nutrients necessary for human health, including fat-soluble vitamins, energy and bioactive lipids. As well, some components have been associated with weight loss and properties that prevent cancer.

Currently, a typical cow's butterfat contains a relatively low level of unsaturated fatty acids, at 25 to 30 per cent, and a higher proportion of saturated long-chain fatty acids, at 70 to 75 per cent. A fat composition more favourable to human health would be about 30 per cent saturated fats and 70 per cent unsaturated fats. Of the FAs in butterfat, the shorter chain FAs (C6-12) are considered a neutral benefit, the longer chain saturated fatty acids (CI4-CI6) are considered a negative and the unsaturated C18 fatty acids are considered positive, benefitting human health. CLA, with possible cancer fighting properties, is a member of the C18 unsaturated group.

Technology is available now to select for certain butterfat traits

Image of a milking machine on an udder

Market attention has become focused on improving health aspects of milk and dairy products. Along with changing the relative proportions of saturated and unsaturated fats, increasing the levels of components such as CLA is possible.

A group of Dutch researchers set out to estimate the genetic parameters of the major fatty acids in milk using samples from nearly 2,000 cows. They estimated heritabilities for 16 fatty acids, about 89 per cent of the total butterfat.

Heritabilities for fatty acids from C4:0 to C16:0 were high, from 0.42 to 0.71. Saturated and unsaturated C18 were lower at 0.25, although still significant. The CLA portion had a higher heritability of 0.42. To put these heritabilities in perspective, those ranked at the high end would be in the same range as growth rate and body size. The heritability for the desirable C18 FAs is considered moderate, but still in the same order as milk production, which has responded well to genetic selection over many years.

The researchers looked at the proportions of the different groups of FAs. Short-chain fatty acids (C4:0 to C12:0) averaged 15 per cent of the fat, medium chain (CI4:0 to CI6:0) accounted for 44 per cent and C18 averaged eight per cent. The ratio of saturated FAs to unsaturated was 2.8:1, or 74 per cent saturated and 26 per cent unsaturated.

They looked at variation due to genetic effects versus herd effects. Herd effects were generally smaller, at 0.25, for most of the FAs, except the unsaturated C18, which had larger herd effects of 0.50.

Some differences in genetic effects in individual FAs come from the manner in which they are formed. The cow forms short-chain FAs, and long-chain FAs originate from plant materials she eats. We can increase the concentration of long-chain, unsaturated FAs by altering the cow's diet. This can be done with DHA or the omega fatty acids, and to a certain extent with CLA.

Changes due to feeding get results quickly and effectively. Genetic gain, although slower, is essentially permanent and cumulative for as long as we practice selection to improve that trait. We may achieve short-term gains through feeding, but need genetic gain to make a difference over time.

Genetic correlations indicated that selection for fat percentage overall has little effect on the short-chain fatty acids. However, it has a positive effect on medium C14-16, which are the highest proportion, and a negative effect on the unsaturated C18, including the CLA. Considering most countries have ongoing selection for increased fat and fat percentage, these correlations have some significant implications. By selecting for increased fat percentage, according to these correlations, we are taking milk composition in the wrong direction.

Genetic gain, although slower, is essentially permanent and cumulative

The example used in the Dutch research is that fat percentage increased to 4.4 per cent in 2005 from 3.7 per cent in 1950. Increased fat percentage would lead to the expectation there would be more saturated C16:0 and less unsaturated C18 FA. When comparing 1974 data to the present numbers, the study found this relationship to be true: as fat percentage went to 4.36 per cent from 4 per cent, C16:0 went to 32.6 per cent from 25.5 per cent, and C18 dropped to 21.6 per cent from 31.1 per cent.

It is reasonable to expect a similar situation has developed in Canada, although perhaps not the same degree. The Netherlands has practised a more extreme selection for fat percentage than Canada.

Limited studies on individual FAs have been conducted for the past 30 years, but the information has not been used or applied. This is mainly because gas chromatograph assay testing, used to analyse FAs in butterfat for those studies, is expensive and complex.

A much cheaper, easier method is needed to give producers routine FA test information on their cows. Belgian research published in 2006 suggests such a method.

The Belgians tested for FA levels using the Foss T6000 model for midinfrared measurement. This is essentially the new model of infrared spectrometer commonly used in milktesting laboratories, such as DHI labs, around the world. With some calibration, this machine could measure most FA fractions in butterfat.

Labs that have upgraded one or more testing machines have a few T6000s. However, the new infrared testing equipment is not found in every milk-testing lab today. This equipment is expensive and replaced infrequently.

Upgrading our labs to new equipment would be costly. On the other hand, installation of one new machine, along with the necessary calibration trials and testing to allow running at least a portion of routine samples for FAs, appears to be within reach. Testing could be done on milk samples as they go through the DHI system. It would make FA testing possible, unlike the use of chemical assays, which are totally impractical for widespread use.

To decide whether a selection program would be successful, three critical questions need to be answered:

  • Can the trait be measured practically?
  • Is there adequate genetic variation?
  • Is there adequate economic reward?

The two research reports appear to answer the first two questions about genetic variation, and the possibility of widespread testing. Will there be an economic reward for producers to improve the FA make up of butterfat? The industry needs to address that Issue.

The results of these two sets of research provide some exciting possibilities for the future of composition of milk and butterfat components. They also provide some challenges. If we can identify cows and family lines for high components of unsaturated fats or CLA, how will producers get rewarded for producing this milk? Will this be treated as specialty milk which would have to be trucked and processed separately, just as DHA milk is now? On an industry basis in order to move milk fat composition to a more favorable mix of desirable FA, producers could be paid per kg produced or on a differential just as payment is made for fat and protein components now.

This research definitely merits a second look for its application in Canada. Purchase of new testing equipment would facilitate identification of superior cows, and possibly selection for different FA composition. Identification of individual animals for specific FAs such as CLA would also be possible. Selection could have great benefits in human health through improving butterfat's health qualities.


Stoop, W.M., J.A.M. van Arendonk, J.M.L. Heck, H.J.F. van Valenberg, and H. Bovenhuis. 2008. Genetic parameters for Major Milk Fatty Acids and Milk Production Traits of Dutch Holstein Friesians. J. Dairy Sci. 91:385-394.

Soyeurt, H, P. Dardenne, F. Dehareng, G. Lognay, D. Veselko, S. M. Marlier, C. Bertozzi, P. Mayeres, and N. Gengler. 2006. Estimating Fatty Acid Content of Cow Milk Using Mid-infrared Spectrometry. J. Dairy Sci. 89:3690-3695.

This article first appeared in the Ruminations column of The Milk Producer Magazine, February 2008.

For more information:
Toll Free: 1-877-424-1300
Author: Blair Murray - Dairy Genetic Improvement Specialist/OMAFRA
Creation Date: 01 May 2011
Last Reviewed: 01 May 2011