Paratuberculosis (PTB) is a fatal chronic intestinal infection caused by Mycobacterium avium subsp. paratuberculosis (MAP), affecting primarily domestic and wild ruminants such as red deer, but also other wildlife including rabbits, foxes and badgers (1–4). Also known as Johne’s disease (JD), PTB is characterized by a long incubation period during which animals may remain subclinically infected for years (5). The cardinal clinical signs, weight loss and watery diarrhea, develop at a late stage of infection; finally, the disease leads to the death of affected animals. Young animals under 6 months of age are at the highest risk of becoming infected in herds with PTB (6). First reported in Germany at the end of the 19th century, PTB is now widely distributed throughout the world and is considered a significant disease due to its multiple negative impacts on economy and animal welfare (5, 7–9).

Intensive animal purchase, especially from multiple herds of origin, is a major risk factor associated with MAP introduction into cattle herds (10–13). Contact with feces of infected animals is also a potential risk factor for the introduction of the disease, and herd size is a recognized risk factor associated with a positive herd status as well (14, 15). Within-herd MAP transmission is primarily associated with contact of calves with adult cows’ feces (16), whereby transmission to young calves occurs mainly by ingestion of milk or feed contaminated with fecal material from infected animals (5). Beside indirect contamination of colostrum and milk through fecal material, MAP may also be directly secreted by the mammary gland of infected cows, especially in late stages of the disease (17–19). Intrauterine infection has also been described, particularly in cows with clinical JD (20). Management of the calving area and contact of the newborn calves with their dams are important points in the control of within-herd transmission (21, 22).

Mycobacterium avium subsp. paratuberculosis has been suspected to represent a zoonotic risk as a trigger agent for Crohn’s Disease (CD), a human chronic inflammatory bowel disease (23–29). Since MAP can be secreted directly into the milk of infected cows (18), a possible transmission way to humans could be through consumption of milk and dairy products (30). In many countries, cow milk is generally pasteurized prior to consumption, however, MAP has been shown to be able of surviving commercial pasteurization (31, 32). Insufficient understanding of MAP’s zoonotic potential and of its role in the development of human disease hampers a well-founded evaluation of the magnitude of its impact on public health (33, 34).

Economic losses due to PTB are mainly attributable to reduced milk production, animal replacement costs and decreased slaughter value (9, 35–39). In Switzerland, an annual economic loss of approximately CHF 4.6 millions due to reduced milk production was recently calculated for a population of 559,900 dairy cows (40). The fact that PTB can be spread through the movement of subclinically infected animals that contaminate their new environment, e.g., after purchase in a new herd, is considered to have contributed to the dissemination of JD (e.g., infected animals of continental European origin suspected of being a source of JD in Ireland), and has therefore serious implications for animal trade (9, 13, 41–44).

Although numerous countries have formal, mostly voluntary PTB control programs based on testing and culling strategies (9, 45, 46), these programs mostly have limited success due to the long incubation period and the low efficiency of diagnostic methods to identify infected animals during the subclinical phase of the disease (41, 47, 48). The sensitivity (Se) of different diagnostic tests varies considerably, ranging, e.g., between 7 and 94% for serum Enzyme-Linked Immunosorbent Assays (ELISA), 29 to 61% for milk ELISA, and 23 and 74% for fecal culture (FC), however, variations in study design and used diagnostic methods make comparison of different results almost impossible (49, 50). The delay of several years between infection time and detectable MAP shedding or immune reaction means that the same test is likely to perform better when used for animals with clinical JD than for animals in the subclinical stage of infection, e.g., the Se of a serum ELISA was found to be 87% in cows with clinical PTB and only 15% in subclinically infected animals (51–53). Likewise, serum ELISA showed a higher Se in animals with a heavy bacterial load (>50 colony forming units (CFU) per tube; Se = 75%) compared to low shedders (<10 CFU/tube; Se = 15%) (54). While FC allows for detecting infected animals earlier in the course of disease than other methods, it requires long incubation times (up to 16 weeks) until a definitive result is available (5, 52), and it can only detect animals excreting the organism, resulting in false-negative outcome in infected animals that have not yet started to shed the bacterium (55). Furthermore, the intensive laboratory work and expertise required for culturing are associated with high costs. In contrast, PCR methods are faster, easier and cheaper, but it has been shown that they may have a distinctly lower Se for MAP detection than FC (56). Therefore, despite the disadvantage of limited Se, quick and unexpensive serological tests are still commonly used to determine the prevalence of MAP infection at the herd level (49).

Numerous studies have been conducted worldwide to estimate PTB prevalence in cattle (9, 15, 46, 57, 58). More than 20% of the herds were considered to be infected with MAP in approximately half of 48 countries around the world for which data were available (9). The disease has been recently classified as enzootic in 27 European countries (including Switzerland), i.e., “countries where the disease was present and for which all periods of absence were shorter than 2 years” (46). Prevalence estimates have generally been higher for dairy cattle than for beef cattle (7). In North America, a study conducted in the 17 main dairy U.S. states in 2007 indicated that 68.1% herds had at least one cow that tested positive on FC (59). In Europe, the apparent between-herd seroprevalence in cattle was reported to range from 38 to 68% (60). However, available prevalence studies are heterogeneous and it is difficult to compare their results due to differences in sampling design, diagnostic strategies and case definitions (41). In many regions of the world, the prevalence of PTB is still completely unknown (45).

In Switzerland, where PTB is a notifiable disease, approximately 60 bovine cases have been reported yearly in the last 3 years (61), however, given the long subclinical phase of infection and unspecific clinical signs, PTB may be more common but affected animals are culled without diagnosis confirmation. Sparse prevalence data at the herd and at the animal level are also available from earlier studies (62–66), but current information about the actual prevalence of PTB in Swiss dairy herds is not available, despite the importance of the dairy industry in Switzerland. The aim of the present study was to determine the between-herd and the within-herd seroprevalence of PTB in a representative subset of Swiss dairy herds. A serum ELISA method was used to allow for inclusion of large numbers of animals and herds. In addition, information about the herds was collected in order to investigate risk factors associated with positive herd status and seroprevalence within positive herds.

The main aim of the present study was to estimate the national PTB herd and animal level prevalences in a large sample of Swiss dairy cattle herds, and to evaluate risk factors potentially associated with a positive herd serostatus and with the within-herd seroprevalence in positive herds. The apparent between-herd seroprevalence of 5.5% detected in this survey is distinctly lower than the apparent between-herd seroprevalence reported in previous studies, e.g., in a review of several studies conducted in European countries (38–68%) (60), or in a study conducted in Northern Italy (48% in Lombardy and 65% in Veneto (82)). At the individual animal level, the apparent seroprevalence of 0.53% is also lower than reported among adult cows in other Europeans countries, e.g., 2.6% in Lombardy and 4.0% in Veneto (82) or from 4.4 to 10.3% in Hungary (83). The median apparent within-herd seroprevalence was 3.6% in the nine seropositive herds. Previous prevalence records of MAP infection in Swiss dairy cattle are limited to few earlier studies (62–66). However, none of these studies provided reliable data for Switzerland due to small sample sizes or lack of representativity of the study population. Because participation in the present study was voluntary, a selection bias with farmers knowing or suspecting the presence of PTB in their herd and choosing not to participate for fear of possible consequences cannot be excluded. Although positive serological results had no regulatory consequences for the participating farms (this was confirmed in the mail sent to recruit participants), some farmers may have preferred not to be involved in a study on a notifiable disease. The recruitment of participating herds turned out to be more challenging than expected, although it was done in collaboration with the main breeding associations, with written confirmation by their boards that they recommended participation to this study on a disease of importance to the dairy industry. This eventually led to a number of participating herds lower than the calculated sample size. This suggests that (mandatory) sampling in a randomly selected subpopulation of dairy farms would be necessary for a really representative study on PTB in Switzerland. The exclusion of small farms in order to mitigate limited test accuracy through recruitment of herds with 25 cows or more may have further complicated the acquisition of participants for the study, given the relatively small average size of Swiss dairy herds. However, 40 animals or more were tested in most participating herds (121 out of 163, 74.2%). The farms included in the study were distributed in all agricultural zones of Switzerland (midland, hill, and mountain zones I-IV). This division into agricultural zones aims at representing the degree of difficulty in production and living conditions to be taken into consideration in the application of the Agriculture Act, e.g., some subsidies provided to the farmers depend on the zone where their farm is located. In the study population, a larger proportion of the farms (52.7%) was located in the midland zone compared to the proportion in the entire Swiss dairy population (32.5% of the herds). In contrast, the four mountain zones together were less represented in the study population in comparison with the general Swiss dairy farm distribution (32% vs. 52.2%). The distribution in the hill zone was the same in both populations (15.3%). This observation is likely due to the selection of farms with a minimal herd size of 25 dairy cows, as farms in the mountain zones tend to be smaller compared to the farms in the midland zone, with an average of 30.5 cows and 18.5 cows, respectively, in 2022 (84).

The observed seroprevalence is surprisingly low, despite the large number of samples collected in more than 160 herds. The serological analyses were conducted in duplicate in a certified laboratory (the Swiss reference laboratory for mycobacterial diseases in animals) with a commercial ELISA-kit, thus an unnoticed technical problem appears unplausible. One concern about using ELISA methods for the detection of positive herds is the possible misclassification of farms due to the low Se of serological tests in early stages of the infection, as antibodies are rarely present in sufficient quantities to be detected during the subclinical phase of PTB (21, 51). Based on the characteristics of the test used (Se = 58.2%, Sp = 99.3%), false negative results must be expected at the individual animal level. Given the contagious nature of PTB, it is not surprising that the odds of herd seropositivity should increase with herd size, which has been observed in numerous studies worldwide (14, 15, 66, 74, 85), including serological studies (12, 21, 22, 86). However, the higher probability of finding positive animals when more animals are tested in larger herds than in small herds must also be taken into consideration. This is the reason why we chose to sample and test all animals (from the age of 1 year old) in the participating herds, in order to increase the likelihood of detecting seropositive herds. Indeed, 146 (120 cows and 26 heifers) and 171 (138 cows and 33 heifers) animals were tested in the two farms with 8 and 4 seropositive animals, respectively, between 41 and 76 animals in the herds with two seropositive animals, and 36 in the smallest farm that was classified as positive with one seropositive cow. In order not to mitigate better Se with decreased Sp at the herd level due to the large number of tested animals, a cutoff (minimal number of seropositive animals needed to classify a herd as positive depending on the number of tested animals) was determined for each herd individually to ensure a Se and a Sp of 95% or more at herd level. Herd size, calculated based on the number of cows only in the models in order to include the herds with heifer rearing outside of the farm, remained significantly associated with a positive herd status, indicating that other factors than the number of tested animals alone contribute to the risk of being PTB positive in larger herds. These other factors may include, e.g., more animal purchase, larger groups of animals, or less control of management practices on large farms with hired employees than in small family farms.

The predominant breed in the participating farms was Holstein with 47.9% of the herds. These were mostly large herds, with a median herd size of 46.5 cows. In contrast, Swiss dairy farms are generally smaller with an average herd size of 23 cows (84). Although the Jersey breed has been reported to be more susceptible for PTB than other dairy breeds (85, 87, 88), breed was not identified as a risk factor in our analyses. The Jersey breed is not very common in Switzerland, as only 5,108 cows were registered with the breeding association in 2022 (71), which corresponded to 0.9% of the total dairy cow population in Switzerland (84). In our study, Jersey cattle was predominant in four herds (2.4%), one of which was composed exclusively of Jerseys. Nevertheless, 36% (9/25) of the seropositive animals in the positive herds were Jerseys, and the herd with the highest number of seropositive animals (8/146, 5.5%) was the one with Jersey cattle only. However, given the low number of observations, these results should not be overinterpreted.

Despite the precautions taken to minimize erroneous classification of herds, we know of at least one herd that was likely falsely classified as seronegative in the study as one cow (tested negative by ELISA in November 2021) later exhibited clinical signs of JD; according to the owner, the diagnosis was confirmed by examination of a fecal sample (method unknown). Thirty-five animals (24 cows and 11 heifers) had been tested in the herd, the positive cow was, to the best of our knowledge, the only animal that later exhibited signs of JD. There had already been a case of PTB in that herd in 2010 (confirmed by an unidentified laboratory analysis), but no other case had occurred since then until 2022. This underlines the limitations of serum ELISA at a single occasion even if all animals in the herd are tested. The use of more than one test, repeated over time to establish the disease status both of animals and herds, is recommended due to the well known diagnostic limitations for PTB (89). In the present study, every animal could be tested only once with an ELISA test due to practical constraints.

A secondary aim of the study was to identify the subpopulation at highest risk within seropositive herds, i.e., the one that should be preferably tested to determine the PTB status of a herd in the frame of a targeted sampling strategy. This objective could not be achieved given the low proportion of animals with a positive test result in any herd, as only one animal was seropositive in most of them (65.6%). Specific analyses to explore possible associations between individual serostatus and age or fertility (calving interval) of the tested animals did not reveal significant associations (data not shown). Thus, the results of this study suggest that the whole herd should be tested to establish its PTB status. A recommendation as to whether heifers between the age of 1 and 2 years should be tested to determine the PTB status of a herd cannot be made based on the present results, as none of the heifers that tested positive was in a herd that was eventually classified as seropositive. Nevertheless, positive FC results for MAP were found in a surprisingly high proportion of fecal samples from heifers in a previous study in Switzerland, suggesting that younger animals may play a role in the spread of MAP in infected herds and thus should be taken into account when trying to establish a herd status regarding PTB (66). Repeated testing would likely be the key to improve result accuracy (90). Testing of bulk tank milk using commercial ELISA methods is inexpensive and has been investigated for herd screening for JD, however, the results appear to be influenced by herd size and within-herd prevalence (91). Milk ELISA has been proven not to be sensitive enough to detect low prevalence herds (92, 93), thus it does not appear to be an adequate option to identify PTB infected Swiss dairy herds.

The strongest association between a risk factor and seropositive status of the herd was observed for “contact of lactating cows with cows from other herds during grazing” in the multivariable analysis. The lactating cows of three herds (one seropositive and two seronegative) shared grazing pastures around the farm (not alpine pastures) with cows from other herds. In two herds (herd size of 34 and 16 cows, respectively), the cows shared the pasture with cows from two other herds, in the third herd (44 cows) the cows shared the pasture with cows from nine other herds (data not shown). This practice appeared to be a potential risk factor for the herd of testing positive, however, this result must be interpreted with caution given the low number of farms with this characteristic. More data would be needed to further explore this potential risk factor. Nevertheless, keeping manure from cattle from other farms away from the herd by avoiding community or shared pasture has been recommended to decrease the risk of MAP introduction in dairy operations (94). Common pasturing has rather been suggested as a risk factor for young animals (calves and heifers) sharing pastures with adult cattle (95), which is in line with young animals being most susceptible to infection with MAP (6). In the present study, however, the variable “heifers sharing alpine pastures with animals from other farms” (mostly young animals, in 67.5% of the cases, but also with adult animals in 15.3%) was found to be associated with decreased odds of the herd testing seropositive, which was unexpected, especially as MAP shedding in heifers has been reported in a previous study in Switzerland (66). Common alpine pasturing of young animals is widespread in Switzerland (96) and the importance of alpine communal pasturing for the spread of other infectious diseases such as Bovine Virus Diarrhea (BVD) is well documented (97). In our study, heifers sharing alpine pastures with animals from other herds, was reported in 5 of 9 seropositive herds (55.6%, only with young animals) and 130 of 154 seronegative herds (84.4%, of which 68.2% with contact to young animals only and 16.2% to adult cattle), thus potential exposure to MAP may not have been the same in both groups. Although it may be postulated that the significant association may be related to low contact intensity due to extensive pasturing on alpine pastures or to the fact that PTB is not as highly contagious as BVD, these results must, again, be interpreted with caution given the low number of positive farms. Indeed, the odds ratios of all variables remaining in the final multivariable model exhibit wide confidence intervals (Table 7), indicating substantial uncertainty. Consequently, the interpretability of these risk factors is limited. This uncertainty primarily stems from the small number of seropositive herds and risk factors with limited discriminative power due to the presence of only a few observations within specific groups.

The identification of significant risk factors associated with high within-herd prevalence was rendered difficult by the low sensitivity of the ELISA test and the low within-herd prevalence, therefore the results of the analyses conducted to identify such factors must be interpreted with even greater caution than associations with the serostatus of the herds. It is, indeed, difficult to propose biologically rational explanations for the fact that the occurrence of diseases (mostly claw and udder problems were mentioned by the farmers) at the herd level in lactating cows would contribute to decreasing the risk of MAP spread in a positive farm. It must be pointed out that, e.g., relevant diseases in cows at the herd level were not observed in any of the 9 seropositive herds. Likewise, the calves were fed milk powder only (no fresh milk) in 22.2% of the seropositive herds (2/9) but in only 7.1% of the seronegative herds (11/154); this random repartition in our study population might explain, at least in part, why the feeding of milk powder, a common recommendation for PTB positive herds, appeared to be related to a low risk of MAP spreading.

In general, farmers’ knowledge about PTB was low, with more than 75.5% of the participants unaware of the very existence of the disease, of which 2.4% turned out having a seropositive herd. This suggest that signs of clinical disease would not be recognized as suspect of PTB by these farmers. After fulfilling the questionnaire, 17.2% of the participants could not exclude that they had had cases of PTB in their herd in the past, while most of them (74.8%) still stated that they had never had a case, although 2.4% of their herds were eventually classified as seropositive. Thirteen of the 163 farmers indicated that they had had at least one case (between one and five diseased animals per farm) in the 10 years prior to the study. While almost all 40 farmers who were aware of the existence of PTB were able to mention the cardinal symptoms of diarrhea (97.5%) and emaciation (75%), only 25% mentioned the reduction in milk yield and 5% also a reduction in fertility. The infection pathways were poorly understood by most, only 35% of the farmers aware of the disease were able to mention at least one infection pathway. The two mostly mentioned pathways were via feces (40%) and intrauterine infection (32.5%). Transmission through milk was rarely known (17.5%) and only 5% of the farmers were aware of possible transmission via colostrum. These results suggest that, despite previous studies in Switzerland (66, 98), PTB is still poorly known by Swiss farmers and further efforts to raise awareness are necessary. It is especially important to raise awareness about transmission through milk, due to the previously mentioned zoonotic potential of MAP, especially in a country with an important tradition in dairy production.

The results of the present study revealed a very low between-herd and within-herd prevalence of MAP seropositivity in Swiss dairy farms, even in large herds. This questions the true epidemiologic relevance of PTB in Switzerland. The low within-herd prevalence and the small size of Swiss dairy herds prevented the identification of a target population to be tested preferably to determine the PTB status of a dairy herd. In combination with the low performance, especially the low sensitivity, of serologic tests for the diagnosis of PTB, these results suggest that alternative methods (e.g., repeated PCR of environmental samples) may be more adequate and should be evaluated for Swiss dairy herds. Difficulties in recruiting farmers willing to participate in the study and their low level of awareness of PTB reveal knowledge gaps and poor understanding about infectious diseases, their transmission and the importance of biosecurity measures. Better information and education of Swiss dairy farmers about PTB and biosecurity in general should be a priority in the near future.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

The animal studies were approved by Amt für Veterinärwesen, 3000 Bern, Switzerland. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent was obtained from the owners for the participation of their animals in this study.

MO: Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation, Writing – original draft, Writing – review & editing. IL: Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing, Conceptualization, Funding acquisition, Resources, Software, Supervision, Validation. JW: Data curation, Formal analysis, Methodology, Software, Supervision, Writing – original draft, Writing – review & editing. SS: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Supervision, Validation, Writing – original draft, Writing – review & editing. MS: Formal analysis, Investigation, Methodology, Supervision, Writing – original draft, Writing – review & editing. RSc: Investigation, Writing – original draft, Writing – review & editing. RSt: Conceptualization, Funding acquisition, Methodology, Supervision, Validation, Writing – original draft, Writing – review & editing. MM: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.