Thomas W.M. Nuhfer | July 1, 2026
A few years ago, I started noticing an interesting trend. I had several dairy-intolerant friends who traveled to Iceland and came back reporting that although they couldn’t stomach American dairy, they found Icelandic dairy much more agreeable. A couple friends from India had similar experiences with Indian milk vs. US milk. Some of my friends speculated that this may have been related to differences in cattle diets, medication/antibiotic regimes, or dairy processing. It could even be a placebo effect! But some digestive differences may be attributable to the genetic histories of the cows themselves.
As of 2019, 83.3% of dairy cattle in the United States were Holsteins [1] (Fig. 1). This breed originated from Dutch Friesian cattle, and all US Holsteins are descended from a handful of cows introduced to the US in the early 1800s [2].
Fig. 1 – Black and white Holstein cow (Source: Verum via Wikimedia Commons)
But while Holsteins dominate US commercial milk markets, there is a lot of global variation in dairy cattle breeds. For example, Iceland has only one dairy breed, the Icelandic breed (Fig. 2).
Fig. 2 – Icelandic cow. (Source: Christian Bickel via Wikimedia Commons)
Genetic differences in cattle can go even deeper than breed. Cattle (both dairy and beef) experienced two separate domestication events (Fig. 3) [3, 4]. Taurine cattle (which are now common in temperate areas including much of the US and Europe) were domesticated from wild, now extinct ancient bovines called aurochs in the Black Sea region. Indicine cattle, also called Zebu cattle (Fig. 4), were domesticated separately in South Asia and formed their own subspecies.
Fig. 3 – A map depicting the site of domestication of Taurine cattle (yellow spot) and the more recent Indicine domestication (red spot), as well as successive migration of the lineages into Europe and Africa. (Source: Bitew et al., 2026 [4])
Zebu cattle are more common in tropical regions. Holsteins, and most US cattle, are taurine. Brahman cattle are an example of a common indicine (Zebu) breed. Zebu cattle are better adapted for hot conditions, so there are some crossbreeding efforts to create Holstein x Brahman crosses for dairy production in hot regions of the US like the desert southwest.
Fig. 4 – Brahman (Zebu) cow. (Source: Strawberry Fields Forever via Wikimedia Commons)
But what does this all mean for milk? A cow’s genetic lineage influences the nutritional composition of its milk. Different individuals and different breeds have more or less sugars, milkfats, and protein, even with the same diet. The primary proteins in milk are beta caseins, and they come in two primary forms: A1 and A2. These particular proteins are determined by a certain gene, which can be expressed as A1/A1, A1/A2, or A2/A2 [5] (Fig. 5). The A1 and A2 alleles are co-dominant, thus heterozygous individuals (who inherit an A1 allele from one parent and an A2 allele from the other) have a mixed genotype. In this case, they make mixed milk! A1 and A2 proteins in milk are very similar. However, due to a little mutation, A1 proteins result in the production of a peptide called BCM-7, while A2 proteins do not.
Fig. 5 – A) An example punnet square with an A1/A2 dam and an A1/A2 sire. Generally, 25% of their offspring would be A1/A1, 25% would be A2/A2, and 50% would be A1/A2. B) Each combination of alleles manifests as differences in milk composition. A1/A1 cattle produce milk with only A1 proteins. A1/A2 cattle produce milk with both proteins. A2/A2 cattle produce milk with only A2 proteins. (Source: Thomas Nuhfer, created with Canva)
Some people believe that the BCM-7 peptide can lead to indigestion, though these claims have not been evaluated by the FDA [5] or rigorously tested via human trials. Despite the need for more scientific testing, some companies have noticed a marketing opportunity and started labeling A2 dairy products (made by A2/A2 cattle, they have no A1 proteins and therefore no BCM-7 peptide) (Fig. 5).
Fig. 6 – A2 branded milk. (Source: BlackCab via Wikimedia Commons)
Is A2 milk really easier to digest? The jury is out— and health claims like these should go through very thorough evaluation. Either way, it won’t make a difference for people who are truly lactose intolerant, since lactose is a separate protein. But if you’ve found that milk in Iceland or India goes down easier than milk in the USA, you might consider giving A2 milk a shot. About 35-50% of Holsteins have the A1 allele [6] though that number is likely shrinking [7]. A smaller 32% percent of Icelandic cattle [8] and only 19% percent of Indian Zebu have the A1 allele [9]. And you don’t need to travel— less than 5% of US Guernseys have it [6], so if you can find a farm that milks Guernseys and sells dairy directly (rather than going through a dairy processor), you can test it out yourself. Genetic tests let us identify A2/A2 individuals, and the trait is being preferentially bred in some dairy populations (like Holsteins), so US milk may eventually have more A2 protein and less BCM-7. However, artificial selection for traits like these can come at the expense of genetic diversity. A study of almost 700,000 Holsteins found the genetic diversity to be equivalent to 43 individuals, making the breed very vulnerable to genetic consequences of inbreeding [10] (like harmful genetic conditions, infertility, and disease susceptibility) . If that spoils your appetite— or if your tummy still grumbles after A2 dairy— there’s always oat milk!
References:
[1] Guinan, Fiona Louise, Howard Duane Norman, and João Walter Dürr. “Changes Occurring in the Breed Composition of U.S. Dairy Herds.” Interbull Bulletin, no. 55 (October 2019): 11–16.
[2] History of the Holstein Breed, Holstein Association USA. https://www.holsteinusa.com/holstein_breed/breedhistory.html
[3] Loftus, R. T., D. E. MacHugh, D. G. Bradley, P. M. Sharp, and P. Cunningham. “Evidence for Two Independent Domestications of Cattle.” Proceedings of the National Academy of Sciences 91, no. 7 (1994): 2757–61. https://doi.org/10.1073/pnas.91.7.2757.
[4] Bitew, Mulusew Kassa, Helina Solomon Woldekiros, Christian Persichilli, et al. “Unscrambling the History of African Indicine Cattle Genomes.” iScience 29, no. 4 (2026): 115295. https://doi.org/10.1016/j.isci.2026.115295.
[5] Borş, Alina, Silviu-Ionuț Borş, and Viorel-Cezar Floriștean. “Health-Related Outcomes and Molecular Methods for the Characterization of A1 and A2 Cow’s Milk: Review and Update.” Veterinary Sciences 11, no. 4 (2024): 172. https://doi.org/10.3390/vetsci11040172.
[6] Cieślińska, Anna, Ewa Fiedorowicz, Dominika Rozmus, Edyta Sienkiewicz-Szłapka, Beata Jarmołowska, and Stanisław Kamiński. “Does a Little Difference Make a Big Difference? Bovine β-Casein A1 and A2 Variants and Human Health—An Update.” International Journal of Molecular Sciences 23, no. 24 (2022): 15637. https://doi.org/10.3390/ijms232415637.
[7] Scott, Beth Anna, Mekonnen Haile-Mariam, Iona M. MacLeod, Ruidong Xiang, and Jennie E. Pryce. “Evaluating the Potential Impact of Selection for the A2 Milk Allele on Inbreeding and Performance in Australian Holstein Cattle.” Frontiers in Animal Science 4 (April 2023). https://doi.org/10.3389/fanim.2023.1142673.
[8] Lien, S., J. Kantanen, 1i. Olsaker, et al. “Comparison of Milk Protein Allele Frequencies in Nordic Cattle Breeds.” Animal Genetics 30, no. 2 (1999): 85–91. https://doi.org/10.1046/j.1365-2052.1999.00434.x.
[9] Khan, Rupali, Sachinandan De, Rahul Dewangan, Rajendra Tamboli, and Reeshu Gupta. “Potential Status of A1 and A2 Variants of Bovine Beta-Casein Gene in Milk Samples of Indian Cattle Breeds.” Animal Biotechnology 34, no. 9 (2023): 4878–84. https://doi.org/10.1080/10495398.2023.2200502.
[10] Makanjuola, Bayode O., Filippo Miglior, Emhimad A. Abdalla, Christian Maltecca, Flavio S. Schenkel, and Christine F. Baes. “Effect of Genomic Selection on Rate of Inbreeding and Coancestry and Effective Population Size of Holstein and Jersey Cattle Populations.” Journal of Dairy Science 103, no. 6 (2020): 5183–99. https://doi.org/10.3168/jds.2019-18013.