Infection and inflammation involve altered redox balance and induce oxidative stress (27). Mastitis is associated with the release of free radicals, increased total oxidant capacity, and decreased total antioxidant capacity in milk (28–31). As part of the immune response to udder infection, reactive oxygen species (ROS) are released by immune cells and play a vital role in combating pathogens. Under normal physiological conditions, ROS production and clearance are in dynamic equilibrium, and ROS can be maintained at extremely low beneficial and harmless levels (32). However, an imbalance in the redox status of the individual animal due to excessive ROS production or impaired antioxidant defense can result in oxidative stress. Cows with mastitis compared to healthy cows appear to have approximately a 10-fold oxidative stress index (33). ROS are highly reactive, including hydrogen peroxide, singlet oxygen, and radicals and can interact with lipids and proteins. Increased ROS levels not only decrease milk quality (28, 34, 35) but could also pose a risk to the consumer (36). Lipid oxidation in food produces as primary products lipid hydroperoxides, which then decompose into a variety of oxygenated and aliphatic fatty acid scission products (37). Lipid hydroperoxides are unstable and do not seem to be absorbed into the plasma. Research on the toxicity of lipid oxidation products in foods has therefore focused on secondary lipid oxidation products (37). Specific examples of lipid oxidation compounds in food that are of health concern include malonaldehyde (MDA) and several cholesterol oxidation products (37–39). As judged from the results of animal experiments, the involvement of oxysterols in atherogenesis is highly controversial because they are also formed in the body and are also involved in some physiological processes but cannot be excluded (40). Cholesterol is up to 10 times less susceptible to oxidation compared to polyunsaturated fatty acids (37).

MDA of approximately 20, 30, and 55 nmol/ml in milk from healthy cows, compared to cows with subclinical and clinical mastitis, respectively, was reported (41, 42). In another study, MDA values were 16 and 46 nmol/ml in milk from either healthy cows or cows with subclinical mastitis (43) suggesting up to triplicate MDA contents when milk is obtained from animals with udder infections. Milk lipid oxidation and MDA contents are also affected by the level of production (44). Milk from healthy cows in their second lactation, at peak lactation, had increasing MDA levels of approximately 20, 40, and 60 nmol/ml at daily milk production levels of 37, 45–49, and 50–56 kg, respectively. This indicates that milk MDA levels at higher production levels with mastitis could be further altered compared to non-mastitis conditions. Subclinical mastitis reduces milk production (45, 46), but higher production levels still may be reached. Mastitis milk contains higher cholesterol levels (47) suggesting also higher contents of cholesterol oxidation products. As mentioned before, also increased protein oxidation may occur with udder infections. However, so far, specific measurements of cholesterol, protein, or other lipid oxidation products in dairy products associated with mastitis are not reported in the literature.

Radicals, other ROS, and cholesterol oxidation products have also endogenous origin, can have physiological functions, and are combated with antioxidants of both exogenous and endogenous origin, to maintain the body in a balanced redox status (32, 40, 48–50), and the oral toxicity of oxidized lipids seems to be low (51). Various food categories including full-fat milk and cheese were analyzed for MDA and two other lipid oxidation products in combination with consumption data obtained from a national representative sample of the Belgian population, for quantitative exposure, and risk assessment (52) applying the Threshold of Toxicological Concern (TTC) concept for chemicals with no toxicity data available. Based on the Cramer decision tree, MDA was classified into class I considering a TTC level of exposure of 30 μg/kg body weight per day. Under this condition, only a small proportion (3.8%) of those who consume cured and minced raw meat could be at risk. According to that study, the consumption of milk and cheese, however, does not appear to pose a risk. Higher MDA levels in milk and cheese due to udder infections were not considered in that study but would not probably make a difference in the possible outcomes. Nonetheless, to prevent premature conclusions, various aspects should also be considered. Milk and dairy products are part of a diet and as such mastitis infections and altered MDA contents still could make a difference in whether the entire diet is healthy or not, even if these foods didn't represent a risk as individual ingredients. Furthermore, this analysis assesses only the exposure risk of individual chemical compounds. In the case of mastitis milk and dairy, there are various altered compounds with potential biological activity. Even if these compounds had no affirmed individual effect, their combined presence still could have an effect. Moreover, limitations of the TTC principle include a lack of specificity not accounting for specific endpoints or individual susceptibility. In addition, dietary lipid oxidation products could have a greater impact on individuals with underlying health conditions and, as pointed out below, appear to also target specific organs. Finally, the same authors point out that no robust toxicological data are available yet and that therefore, regardless of the study outcomes, precautions should be taken to prevent lipid oxidation in foods.

Chronic uptake of large amounts of lipid oxidation products increased tumor frequency and incidence of atherosclerosis in animals (51, 53–55). Especially in poorly controlled type 2 diabetic patients, but also in healthy subjects, oxidized cholesterol in the diet was found to be a source of oxidized lipoproteins in human serum (56–58) suggesting that these increased postprandial levels of potentially atherogenic oxidized lipids may contribute to the accelerated atherosclerosis associated with diabetes. Some of the dietary lipid oxidation products, which are absorbed from the gut to the circulatory system, seem to activate an inflammation response that affects not only the circulatory system but also specific organs such as the liver, kidney, lung, and the gut itself (39). As for protein oxidation products, again, the formation of these products would not only compromise milk and dairy quality but may also affect food safety (59). When rats were fed diets containing milk powder varying in protein oxidation state, rats responded with respectively varying degrees of redox state imbalance and oxidative damage in plasma, liver, and brain tissues. Furthermore, hippocampal inflammatory and apoptosis genes were significantly upregulated in groups with higher oxidation states, while learning and memory genes were significantly downregulated (59).

Besides ROS, also reactive nitrogen species are involved in the inflammatory response during mastitis. Nitric oxide-derived oxidative stress caused by mastitis of cows and goats increases milk nitrate (28, 35), nitrite (28, 60), nitric oxide (NO) (28, 35), and S-Nitrosamines (28). Cows with mastitis compared to healthy cows appear to have about doubled plasma and milk NO levels (33, 61). In the acidic environment of the stomach, nitrate can be reduced to nitrite and nitrite converts to NO and nitrosamine compounds. Nitrosamine compounds are formed by the reaction of nitrosatable amines (62). Within these compounds, it may be the nitrosamines that are concerning (63). The International Agency for Research on Cancer (IARC) has considered nitrosamine compounds as probable carcinogens for humans (group 2A) (64). A health risk assessment of nitrosamines was conducted as a pilot study in Iran where 33 healthy adults participated in and completed a food frequency questionnaire for 3 days. According to that survey, and literature data on the concentrations of these compounds in the various food categories, dairy products seemed to have a significant impact on the daily intake of nitrosamines. The highest contributor to the daily intake of nitrosamines was meat, followed by dairy products that corresponded to about a third of that provided by meat. Fruits and grains had a minor impact. The estimated cancer risk from the dietary intake of nitrosamines ranged between 1.74 and 2.22 × 10⁻³ with a 95% confidence interval (62). In this context, mastitis can be a game changer. In goat milk from uninfected udders, S-nitrosamines concentrations were only 0.17 (μM) but increased ~14-fold to 2.4 (μM) in milk from infected udders (28).

Cytokines are peptides that have critical roles in immunology and include interleukins (IL), interferons (IFN), lymphokines, chemokines, tumor necrosis factors (TNF), and transforming growth factors (TGF). They are involved in immune response, health status, and tissue development and can act as tissue hormones in an autocrine and paracrine fashion. Cytokines serve as either proinflammatory or anti-inflammatory agents; on the one hand, as neutralizers of harmful pathogenic effects and on the other hand, as inducers for the maturation and development of the immune system (65). Major depression (66), Alzheimer's disease (67), cancer (68), and autoimmune diseases (69) are accompanied by an inflammatory response. However, it appears that increased levels of proinflammatory cytokines are not just symptoms of these diseases. Recent evidence also indicates that a dysfunctional neuroendocrine-immune interface associated with abnormalities of the “systemic anti-inflammatory feedback” and/or “hyperactivity” of the local proinflammatory factors may play a role in the pathogenesis of atopic/allergic and autoimmune diseases, obesity, depression, and atherosclerosis (70, 71). There are also various studies that prove the role of inflammation in the genesis and proliferation of certain tumors such as lung cancer (72), colitis-associated colorectal cancer, and cholangiocarcinoma (73). It appears from a recent study that used information from a lifestyle questionnaire obtained from 150 healthy individuals and their serum samples that red meat consumption seems to be associated with an inflammatory pattern, characterized by an increase in IL-6 and IL-8. IL-8 levels were also increased with the frequent intake of sweets, while a higher intake of shelled fruits correlated with lower levels of IL-6 (74). High body mass index had a higher impact on the serum levels of C-reactive protein than the consumption of individual food categories and, in that context, the consumption of milk and dairy had no influence. However, the study outcomes may not apply to individuals with pre-existing medical conditions. In addition, this study did not account for possible variations in udder infection status or SCC. As shown in the following paragraph, milk cytokine levels are altered with mastitis. Theoretically, the consumption of foods could contribute to cytokine dysregulation or inflammation not only by stimulating the production of proinflammatory cytokines in the body but also by providing these cytokines directly through the foods.

The leukocytes that come with milk from infected udders are involved in the production of milk cytokines (75). Cytokines are also present in milk from healthy udders, but composition and concentrations change with both clinical and subclinical mastitis (16) depending on time after exposure, pathogen species, and severity of the disease (76). Proinflammatory cytokines as indicators of early inflammation such as TNF-α appear to reach a peak in 1–12 h after exposure followed by a gradual drop (77), whereas other cytokines can reach their maximum levels at up to 7 days or even later (76) depending again also on the pathogen species. In naturally acquired mastitis, milk concentrations of IFN-γ reached up to 20 ng/ml, whereas maximum reported levels of IL-6, TNF-α, IL-1β, IL-8, and TGF-α were about 90 ng/ml, 25 ng/ml, 8 ng/ml, 1 ng/ml, and 0.5 ng/ml, respectively, values that were significantly higher than in milk from healthy udders (76). In comparison, the control milk samples had maximum levels of about only 0–0.15 ng/ml of IFN-γ, TNF-α, IL-1β, IL-8, and TGF-α (76).

Milk cytokines are crucial for the development of the newborn's organism and its immune system (78). Cytokines have been also considered for therapeutic use, and in that context, it has been demonstrated that orally administered interferons and cytokines can exert both local and systemic effects (79). TGF-β, for example, when synthesized as an inactive precursor, can be activated during intestinal transit by multiple mechanisms, e.g., by a low pH of 1.5 (65). It is therefore probable that milk cytokines are bioactive after ingestion and digestion. Cytokines have very complex biological functions and can interact with each other. Some can induce the production of others, act synergistically to enhance their effects, inhibit the expression of others, or stimulate the expression of receptor antagonists or receptors of others (76). For example, IL-1 induces the production of IL-1 itself, TNF-α, IL-6, IL-8, and IL-12. In the cases of TNF-α and IL-6, IL-1 can act synergistically to enhance their effects (76). This suggests that cytokines in food could make a difference even if present in apparently small amounts. Almost 30 different cytokines with specific features (65) have been already identified in milk raising the question about their biological immunomodulatory significance as well as their risk for human consumption, which has still to be determined.

Serum amyloid A (SAA) is a precursor in the formation of AA fibrils which, in turn, are associated with various, especially neurodegenerative, diseases, such as Alzheimer's. It has been demonstrated in rodent models that AA fibrils can be orally transmitted (80). The AA fibrils cross the gut barrier (81) and can trigger the disease. The oral “infectious dose” of AA fibrils, at <1 μg, is comparable with the infectivity of prions (82), and species barriers can be surmounted (81). Also, cooking temperatures are unable to eliminate their amyloidogenic potential (82). Presumably because of similarities in structure and composition, AA fibrils exhibit similar resistance to physical and chemical decontamination methods. Treatment with cooking (82), freezing/thawing, and disinfectants such as formalin and 2N NaOH were not able to abolish AA fibril infectivity. Not even autoclaving for 3 h could guarantee inactivation (82). Transgenic mice that were fed extracts of fibrillar material composed of serum amyloid from fois gras developed amyloid deposits in virtually all organs examined (80). The authors concluded that an amyloid-containing food product hastened the development of amyloid protein A amyloidosis in a susceptible population. However, AA amyloidosis seems to only develop in the concomitant presence of high levels of SAA with AA-derived fibril seeds (83). Healthy mice exposed orally to AA fibrils did not develop amyloidosis, whereas those additionally receiving a concurrent inflammatory stimulus developed a pronounced disease within days (83). Conditions associated with elevated SAA levels may include chronic infections, as well as non-infectious chronic inflammatory diseases, and certain tumors (84). The concentration of SAA can drastically increase at an acute inflammation as a response to cytokines, especially IL-1, IL-6, and TNF-α (85). On the other hand, it also has been shown that even months after fibril exposure, an inflammatory stimulus could rapidly induce AA amyloidosis to the same extent as concurrent inflammation and fibril injection (83). In mice, the longest interval between exposure and inflammation studied was 180 days, which corresponds to nearly one-quarter of the animals' natural life span (83). This would suggest that consumers of food with high levels of AA fibrils may be at increased risk for AA amyloidosis in case they developed an inflammatory disorder even years after consumption.

As discussed before, the milk from animals with mastitis would provide cytokines IL-1, IL-6, and TNF-α. In addition, in response to mastitis inflammation, higher levels of milk AA are expected (86–88), levels depending on the pathogen (87). Compared to pathogen-free milk with an AA concentration of approximately 12 μg/ml, AA levels in milk from animals with subclinical mastitis from Staphylococcus aureus, Streptococcus agalactiae, Streptococcus dysgalactiae, Streptococcus uberis, and Candida spp. were ~42, 60, 72, 92, and 102 μg/ml, respectively (87). Milk from animals with coagulase-negative staphylococci subclinical infections instead was not different from the pathogen-free milk. The AA fibril content in milk and dairy from animals with mastitis may be elevated, considering an altered casein composition. Amyloid fibril appears to be formed by αs2- and κ-casein and seems to be inhibited by αs1- and ß-casein (89). The overall casein concentration in milk from animals with mastitis seems to be reduced (90, 91), especially that of β- (92, 93) and αs1-casein (92). At the same time, the contents of κ-casein may be increased (93). In addition to the altered casein composition, milk from infected udders also exhibits greatly increased proteolytic activity (90, 94, 95). Both, plasmin and plasminogen activators are increased when milk is from infected udders (96–98). Moreover, a higher pH in milk from animals with mastitis (90, 99) favors the plasminogen activators (100). Plasminogen activation and plasmin activity appear to be involved in transthyretin amyloidosis including the initiation, progression, and tissue distribution of amyloid deposition (101, 102).

Recently, a concern has been raised that AA amyloid in the human food chain could represent a possible biohazard (103). On that behalf, it was investigated whether amyloid seeds from different food proteins (lysozyme, β-lactoglobulin, soybean, mung bean, fava bean, lupine, potato, oat) affected the kinetics of Aβ₁₋₄₂ amyloid formation, which is the particular variant associated with Alzheimer's disease (104). None of the tested seeds had fibrillation-promoting effects. However, the authors noted differences between the protein sources. Lysozyme, for example, had even an inhibiting effect on Aβ₁₋₄₂ fibrillation. This suggests that the results of this study may not be exhaustive and may not represent all food proteins. Other food proteins, that have not yet been tested, still could promote fibrillation. Amyloid fibril formation by casein and fatty acids was observed in the human breast milk of mastitis patients (105).

In addition to the abovementioned compounds, intramammary infections also increase milk cortisol levels (106). Milk cortisol levels with SCC of <200,000 cells/ml, between 200,000 and 400,000 cells/ml, and of >400,000 cells/ml were 470, 470, and 530 pg/ml, respectively (106). However, this difference appears to be small compared to the cortisol levels that are already present in milk from healthy udders, suggesting that mastitis milk could only worsen to some extent a possible persistent risk. Fludrocortisone, which has over a 100-fold mineralocorticoid potency compared to hydrocortisone, is used to treat adrenocortical deficiency and orthostatic hypotension, and daily doses of 0.1 to 0.2 mg are considered to have only very little or no glucocorticoid effect at all (107). It is rather unlikely to possibly reach similar levels of cortisol from the consumption of milk and dairy products alone but we might also keep in mind that cortisol is a hydrophobic molecule, that should be more concentrated in whole-milk cheese and cream than in milk, and that cortisol may also be present in other animal-derived foods.

Governments and food safety authorities should support and commission research to assess whether any of these bioactive compounds could, considering also their combination, represent any risk to the consumer's health considering also possible long-term effects. We might then also need to further investigate the possible concentrations of these compounds in milk from cows with mastitis. This information would be necessary for the legislators to either confirm that the current regulations are sufficient to guarantee the safety of milk and dairy products or to eventually adapt the SCC thresholds for the commercialization of milk and dairy. However, in any case, when it comes to products with any type of quality label, more restrictive regulations concerning milk hygiene should be considered. These labels not only give assurance about the traceability of the products to specific areas of production and/or the application of specific sets of competencies and know-how, but the consumer may also associate this assurance with distinctively high quality, in general, as well as high sanitary quality and safety standards, in particular, even if these standards are not taken into account.

To reach the goal of milk from mastitis-free animals, regulations alone cannot be sufficient, but financial support for farmers may be necessary. In Europe, mastitis control has become even more difficult since 28 January 2022, when the blanket antibiotic dry cow therapy has been prohibited under Regulation (EU) 2019/6 on veterinary medicinal products. This restriction ought to reduce on-farm antibiotic use to reinforce the EU's strategy against antimicrobial resistance. Selective dry cow therapy is suggested as an alternative instead. However, many farmers will need training, and the latter practice may require also higher attention from the farmer and more specific testing of the individual animals that are to be dried off. Selective dry cow therapy compared to blanket dry cow treatment may increase SCC after calving, especially when cows at dry-off are not treated with an internal teat sealant (108, 109). It is in particular the small-and medium-sized farms that will need subsidies. There are several management practices and protocols to contempt and prevent mastitis, but as a prerequisite, first of all, the animals with mastitis need to be detected. Therefore, frequent individual SCC measurements are obligatory. The best solution to this problem would be the investment in sensing tools that are incorporated into the milking system. In many modern farms of a certain size, such tools are already part of automated milking systems and robots, capable of providing online measurements and immediate alerts to the farmers. Many of the smaller farms instead, where the economic sustainability of dairy farming is at risk, may not be able to afford such sensors, and subsidies should be tailored especially for those situations. With this regard, it should be kept in mind that very often Common Agricultural Policy's subsidies are more likely to reach the larger farms (110). Another economic aid could be provided in certain circumstances for the culling of animals with ascertained specific contagious intramammary infections when it is unlikely that antibiotic treatments can eradicate the infection. Antibiotic treatments of animals with chronic Staphylococcus aureus infections, for example, have a very reduced chance of only 35% to be efficient (111), and increase the development of antibiotic-resistant strains at the same time (112). Animals with contagious mastitis infections should be segregated from healthy animals in the milking parlor, but in the longer term, this practice cannot be able to eliminate the risk of spreading the infections. On the other hand, for the farmer, the decision to cull these animals is often very difficult, especially when large parts of the herd are involved, and the number of replacement heifers is not sufficient to maintain the herd size at a certain level.