Goat Milk Cell Counts Defy Simple Thresholds, Review Finds

Somatic cell count has long served as the workhorse indicator of udder health and milk quality in dairy animals, but a new review argues that in goats the measurement resists the kind of straightforward interpretation that works for cattle. Published in Acta Veterinaria Scandinavica, the analysis synthesizes 119 publications selected from an initial pool of 598 records, covering literature from 2000 to 2025 drawn from Web of Science, Scopus, PubMed and Google Scholar. Its central conclusion is stark: somatic cell count in goat milk cannot be reliably judged against fixed threshold values, because the number reflects a tangle of physiology, genetics, season, management and infection that varies from animal to animal and farm to farm.

The root of the difficulty lies in how goats make milk. Goat mammary secretion is primarily apocrine, meaning that secretory cells release fat globules wrapped in pinched-off pieces of their own cytoplasm. These fragments, usually lacking nuclei and measuring roughly 150,000 per millilitre, are similar in size to somatic cells and can be miscounted as such by routine methods, which is why DNA-based counting techniques are recommended for goats. Sheep milk is also secreted apocrinally, but their cytoplasmic fragment concentration averages only about 15,000 per millilitre, while cows secrete merocrinally, releasing milk components without cellular debris. This mechanistic difference explains why uninfected goats routinely carry somatic cell counts between 270,000 and 2,000,000 cells per millilitre, far above the 10,000 to 200,000 range typical of uninfected sheep and the values of roughly 100,000 considered normal in healthy cows.

Cellular composition adds another layer of complexity. Neutrophils make up 45 to 74 percent of somatic cells in uninfected goat milk, whereas macrophages dominate in cows and sheep. In infected goat glands, neutrophil proportions rise to 71 to 86 percent, macrophages fall to 8 to 18 percent and lymphocytes drop from 9 to 20 percent to 5 to 11 percent, while epithelial cells can account for up to 27 percent of the count in uninfected animals. Because neutrophils are already so abundant in healthy goats, the dramatic shift that signals subclinical mastitis in cattle is muted in goats, weakening one of the classic diagnostic cues. The cells themselves also alter milk chemistry: they release proteolytic enzymes such as plasmin and cathepsin D, and elevated counts are associated with degraded casein, reduced curd firmness and impaired cheese yield.

Regulation reflects the unresolved debate. In the United States, the legal limit for goat milk is 1,000,000 cells per millilitre, with a bulk tank limit of 1,500,000, while sheep milk is held to 750,000. The European Union has no uniform legal threshold for goat milk, and proposals have circulated around 1,500,000 cells per millilitre. The review notes that physiological elevations, particularly in late lactation or during seasonal changes, may not compromise milk safety, whereas inflammation-associated rises can involve pathogenic microorganisms and toxins that damage quality, safety and processing properties. That gap between the two scenarios is precisely what fixed thresholds fail to capture.

The physiological drivers are numerous and sometimes contradictory across studies. Somatic cell count rises with parity up to at least the fifth lactation in some datasets, with the proportion of low-count recordings falling from 45.2 percent in first-parity goats to 20.8 percent in older animals, yet other studies report the opposite trend. Oestrus delivers a transient spike, averaging an increase of 315,000 cells per millilitre and reaching 422,000 in fourth-parity goats, while hormonally induced oestrus can triple counts regardless of infection status. Lactation stage is the most consistent factor, with counts climbing in late lactation, although one study found a decline after 300 days in milk, possibly reflecting selection of favourable animals. Season interacts with the goat’s photoperiod-driven reproductive cycle: reported peaks fall in winter in some regions, in October or June through August in others, with one study attributing 13 percent of count variability to season alone. Even litter size produces conflicting results, with some studies finding higher counts in goats nursing twins and others finding higher counts in goats suckling fewer kids.

Management and environment layer further variability on top. Udder shape matters in some datasets, with pendular cylindrical udders averaging 9,058,100 cells per millilitre against 1,585,300 in pear-shaped udders, but other studies find no consistent relationship with udder dimensions or teat size. Milking type yields contradictory findings as well: one study reported higher counts in machine-milked goats, another in hand-milked herds, and milking frequency shows effects in some trials and none in others. Farm type accounted for 16 percent of variability in one analysis, yet comparisons of conventional, traditional and organic systems found no statistically significant differences despite widely differing mean values. Acute environmental stress raises plasma cortisol and significantly elevates somatic cell counts, particularly in older goats, while cutting milk production by 11 percent on the first day after exposure, demonstrating that stress alone can push counts up without any infection present. Feeding interventions show promise: negative dietary cation-anion difference diets, moringa pellets, purple Napier grass silage, marine algae supplementation at 10 grams per day, and Echinacea-based supplements have all been associated with lower counts in individual trials, though responses depend on supplement type, dose and individual variation.

Genetics contributes a measurable but moderate component. Heritability estimates for milk somatic cell score in Alpine and Saanen goats are low to moderate, varying with breed, parity, lactation stage and statistical model. Specific polymorphisms have been linked to counts, including four single nucleotide polymorphisms in the lactoferrin gene promoter, a C/T variant in exon 4 of the growth hormone receptor gene, and the rs650082970 variant in the TLR2 gene, whose A allele has been suggested to confer increased resistance to mastitis-related inflammation. Breed comparisons are similarly inconsistent: Hair goats showed higher counts than Alpine and Saanen crosses in one study, Saanen goats exceeded several Mediterranean breeds in another, and several studies detected no significant breed differences at all, with substantial individual variability within breeds. The review concludes that genetic factors modulate baseline levels rather than acting as primary determinants, and that selection should treat the cell score within multi-trait indices rather than as an isolated breeding criterion.

Intramammary infection remains the principal pathological driver, but even here the picture is nuanced. Coagulase-negative staphylococci are the most common pathogens, causing persistent subclinical infections, while Staphylococcus aureus, though found in only 3.6 percent of cases in one study, exerted a more pronounced impact on counts than the more frequent Staphylococcus caprae, S. warneri and S. epidermidis. In one dataset, parity-adjusted mean counts were 780,000 cells per millilitre in culture-negative goats and 794,000 in culture-positive goats, compared with 522,000 in goats without bacteriological samples, and even uninfected goats exceeded 1,000,000 cells per millilitre in advanced lactation. Most tellingly, roughly 20 percent of samples with counts below 1,000,000 cells per millilitre still contained pathogens, underscoring the limits of count-based screening alone. Milk electrical conductivity offers a complementary signal, averaging 13.69 millisiemens per centimetre in unhealthy glands versus 12.70 in healthy ones, but automated detection algorithms achieved global sensitivities of only 33.3 to 58.3 percent for clinical and subclinical mastitis, with specificity between 77.8 and 100 percent, and low disease prevalence inflates false positives.

The review’s proposed answer is an integrated interpretation framework that weighs physiological context, infection status, genetic background and modern monitoring tools together rather than relying on any single number. Machine learning models using decision trees, random forests, XGBoost and neural networks have shown strong performance for predicting milk quality parameters on small ruminant farms, and among laboratory methods, direct microscopic counting with May-Grünwald staining proved the most accurate and economical option. The authors acknowledge that heterogeneity among the reviewed studies, including differences in breeds, lactation stages, mastitis definitions and measurement techniques, limits direct comparability and likely explains many of the contradictions. Their recommendation is methodological standardization and context-specific diagnostic criteria, so that a rising cell count in a late-lactating, oestrus-cycling or freshly stressed goat is no longer mistaken for disease, and a genuinely infected gland is not waved away as normal physiology.

Subject of Research: Factors influencing somatic cell count variability in goat milk and implications for udder health assessment

Article Title: Physiological, genetic and inflammatory factors influencing somatic cell count in goat milk: a review

Article References: Einorytė, J., Palubinskas, G., Šlyžienė, B., Baltaduonytė, R., & Šlyžius, E. (2026). Physiological, genetic and inflammatory factors influencing somatic cell count in goat milk: a review. Acta Veterinaria Scandinavica, 68(1), Article 43. https://doi.org/10.1186/s13028-026-00880-7

Image Credits: AI Generated

DOI: 10.1186/s13028-026-00880-7

Keywords: goat milk, somatic cell count, mastitis, udder health, apocrine secretion, intramammary infection, milk quality, genetic polymorphisms, lactation stage, electrical conductivity, dairy goats, machine learning