Bovine mastitis is the most common disease in dairy herds and leads to major economic losses in dairy farms (1). Staphylococcus (S.) aureus is one of the most frequently occurring mastitis pathogens, causes huge costs (2) and can lead to chronical intramammary infections. In a recent prevalence study from Lower Saxony, Germany, S. aureus was detected in bulk milk in 18.3% of the investigated dairy herds (3). Diagnosis is particularly difficult as the pathogen is shed cyclically, and only small quantities of the pathogen are excreted at times (4, 5). Intermittent shedding patterns and the resulting diagnostic challenges have also been investigated in other mastitis pathogens such as Prototheca spp. and Mycoplasma spp. (6, 7). S. aureus has different pathogenic properties to circumvent the immune response of the host (8), e.g., the ability to form a biofilm is suggested to be a major factor (9), although biofilm formation in vivo has rarely been investigated (10). While pathogen quantities of 55,000 cfu/mL are excreted in severe cases of mastitis (11), the excretion of S. aureus in milk is partially below the detection limit of standard laboratory diagnostics (12–14). False negative results due to such intermittent shedding can prevent the successful sanitation of a dairy, therefore research should prioritize on early detection to control the disease through treatment and management (15). The aim of this mini review is to obtain an overview on the occurrence of intermittent S. aureus-shedding in milk and to evaluate the sampling strategies and diagnostic possibilities.

The excretion of S. aureus in milk can differ greatly from day to day and even from two infected quarters of the same cow (5). Sears et al. (4) were able to show that the detection rate of S. aureus could be increased from 75 % to 94 % and respectively 98 % following a second and third sample of the quarter. The time of milk sampling and handling also was found to be relevant: the likelihood of detecting subclinical S. aureus-infection is higher in fresh milk samples taken before milking than in samples taken after milking (16). Villanueva et al. (17) suspected that freezing milk samples could destroy bacterial cell aggregates, which would improve the sensitivity of BC. Such a positive effect could not be confirmed in comparison with fresh pre-milking samples in a study by Godden et al. (16). The freezing of milk samples seems to have different effects on pathogens, whereby no effect could be confirmed especially for S. aureus (18). A study with centrifuged milk samples showed that cultures from the sediment of quarter milk samples can increase the number of positive results by up to 145.5 % (19). Furthermore, Mahmmod et al. (20) demonstrated that pre-sampling procedures (cleaning, disinfecting and discarding first milked streaks) significantly reduced the likelihood of false-positive S. aureus results by eliminating colonies from the skin and teat canal.

Hence, for detecting pathogens in milk different sample types can be used. Quarter milk sample (QMS) are the standard method used for the detection of intramammary infections (IMI) (21). Maisano et al. (22) suggested that the serial sampling of composite milk samples is an alternative to a few QMS. To assess the infection status of a herd, bulk milk samples could be analyzed for the prevalence of S. aureus (23, 24). Britten (13) recommended the targeted use of selective media in general for composite and bulk milk samples to improve sensitivity. However, these results do not allow any statement to be made about the prevalence of the pathogen in individual cows.

In general, enrichment processes are applied before microbial cultures and initiated in specific sampling situations (21). Keefe (14) assumed that the level of S. aureus-excretion may be below the detection limit of conventional BC and stated that enrichment media could be a helpful tool to increase sensitivity in case of low S. aureus-shedding. Also in human medicine for isolation of S. aureus from swab sample selective enrichment broths were used (66, 67). Furthermore, enrichment media for S. aureus were used to investigate the food safety of milk powder and compared with direct plating at selective Baird Parker medium and Hauschild pork plasma fibrinogen medium (68). In this study, the enrichment medium (Giolitti Cantoni broth with Tween 80) did not achieve better results than direct plating on Baird Parker medium but compared to direct plating on Hauschild medium it did (68). Only a few studies have investigated the potential of enrichment broths for more sensitive bovine mastitis diagnostics with different results (30, 69, 70): Thurmond et al. (70) showed in an experiment with composite milk samples that pre-enrichment of milk samples in Brain heart fusion increases the probability of isolating S. aureus by 1.6 times. Artursson et al. (30) could not demonstrate positive effect on the isolation of S. aureus with nutrient broth containing 10% horse serum in analyses of subclinical quarter milk samples. In some studies an enrichment of the milk was done before PCR to carry out more sensitive diagnostics (71–73). Further studies with different enrichment media for the detection of subclinical S. aureus-infections could provide further insights.

Intermittent shedding of S. aureus in bovine milk samples at cow level was investigated in six longitudinal studies shown in Table 2 (4, 5, 12, 33, 36, 39). Despite an extensive literature search from 1980 to 2024, only these six studies provided information on the shedding of S. aureus at cow level, of which five had detailed information on intermittent shedding.

The six studies differ slightly in scope of sample, with 4–22 S. aureus infected cows being analyzed, all examined QMS using BC as detection method. Studer et al. (36) additionally made use of qPCR (see Table 2). Sears et al. (4) investigated naturally and artificially infected cows with S. aureus in an experimental model. In four of the longitudinal studies, the duration of sampling ranged from 6 d to 28 d with at least one sample taken daily. In a study by Walker et al. (5) samples are taken three times on 21 consecutive days during lactation. In another study (33) quarter milk samples are taken weekly over a period of 26–44 weeks. In those studies (5, 33) the bacterial genome was additionally analyzed using pulsed field electrophoresis. Consequently, it should be noted that the six studies differ greatly in terms of study design, which makes comparability difficult.

In three studies, undetectable intervals (see Table 2; “Duration of non-detected shed”) were not investigated in detail, only three studies mention undetectable intervals of ≥0.5 d or 1 d (4, 36, 39). Walker et al. (5) determined that 97.5% of the samples were positive. Unfortunately, the duration of non-detected shed and the number of those cows was not reported. Similar to the study by Buelow et al. (12) in which 11 negative QMS are mentioned, but the number of cows and time interval were not discussed.

Walker et al. (33) compared the results of their study with those of Sears et al. (4) and concluded that naturally infected mammary glands excrete S. aureus in a more consistent pattern than experimentally infected mammary glands. In addition it was found that S. aureus strains have different affinities for the mammary gland, which is why the strain selection might achieve different results (33, 74). However, Sears et al. and Walker et al. (4, 33) agreed on the definition of different types of shedding, so-called low shedding pattern (≤ 10 cfu/0.01 mL) and high shedding pattern (≥20 cfu/0.01 mL). Furthermore, Walker et al. and Studer et al. (33, 36) both observed sinusoidal shedding pattern, i.e., an alternation between low and high S. aureus shedding over time. Studer et al. (36) suggested that this is a result of a synchronized process between the pathogen and the immune system. However, Walker et al. (33) noted that the duration and amplitude of each pattern varied, so that no consistent pattern was observed between or within cows. In a few longitudinal studies transmission of S. aureus was examined on herd level and focusing on the transmission of S. aureus examining the different genotypes (75, 76). Sommerhäuser (75) concluded that the persistence of the pathogen in the udder tissue is more likely in herd sub-suspensions than a temporary cure and subsequent reinfection by the same S. aureus-type. On the other hand Wente et al. (77) investigated recurrent clinical mastitis and stated that S. aureus showed the highest recurrence rate (27 % of all S. aureus cases).

In summary, the occurrence of intermittent shedding of Staphylococcus aureus and the need for sensitive diagnostics is to be investigated in further longitudinal studies. Particular attention should be paid in future research to the duration of undetectable shedding. As this duration was reported as very short in the studies by Sears et al. and Studer et al. (4, 36), close monitoring should also be carried out in future research. Furthermore, whole genome sequencing of the isolates could provide further insights into microevolution in the host in order to determine the persistence of the bacterium in the udder or a re-infection. It is therefore important that a consistent terminology is established to characterize IMI over time (78). Furthermore, there is no standardized definition for the diagnosis of an IMI with S. aureus (32). For BC a sample could already be considered positive if 1 cfu grows in an inoculum of 0.1 mL, as stated by Walker et al. (32, 33). For molecular diagnostics, the subdivision into infected and positive depending on the Ct-value, which was proposed by Cederlöf et al. (51), could be useful. Definitions need to be clarified to enhance research in the field of intermittent shedding of S. aureus in bovine mastitis. Regarding the detection limit, the serological detection and direct MALDI TOF MS studies require further investigations to achieve detection even with low shedding.