Foodborne illnesses pose public health and economic burdens both in developed and developing countries (1, 2). Annually, foodborne illnesses are responsible for an estimated 600 million cases, 420,000 deaths, and 33 million disability adjusted life years lost worldwide. Salmonella is a major cause of foodborne illnesses in humans (3–5). Salmonella are Gram-negative, non-spore forming, mostly motile, facultative anaerobic bacilli within the family Enterobacteriaceae. The species Salmonella enterica consists of six subspecies and more than 2,579 serovars (6, 7). Based on the clinical profiles of infections caused in humans S. enterica can be divided into typhoidal—which are human specific—and non-typhoidal Salmonella (NTS)—having a broad host range (6). The NTS serotypes are leading causes of bacterial diarrhea and invasive bacterial infections in young children, the elderly and the immune-compromised individuals throughout the world. Salmonella Typhimurium and S. Enteritidis together account for approximately 50% of all isolates globally reported from human clinical cases (8–10). The global incidence of diarrheal disease due to the NTS accounts for about 94 million enteric infections each year, of which 80.3 million cases are considered foodborne and resulting in 155,000 human deaths annually (11). Human salmonellosis is also recognized as an important socioeconomic disease posing considerable economic burden in the world (12, 13).

Salmonella colonizes mainly the intestinal tracts of humans and animals including cattle. Foods of animal origin are important sources of Salmonella infections in humans (13–18). Humans acquire the infection mainly through consumption of contaminated products including beef and beef products (19), by direct contact with infected animals or their environment (20) and by direct human-to-human transmission (21). Carcass contamination with Salmonella during slaughter, particularly under unsatisfactory hygienic operations, poses a significant public health risk (22–25). The transfer of NTS to food processing plants and equipment used for food preparation also plays an important role, ultimately leading to the risk of salmonellosis after the consumption of contaminated foods (21). Knowledge about the overall occurrence of Salmonella and the diversity of serotypes in cattle provides important information for decision making and to promote reliable efforts toward prevention and control of foodborne salmonellosis associated with cattle. Therefore, the objectives of this study were to determine the prevalence of Salmonella in apparently healthy cattle, and to assess the diversity of Salmonella serotypes associated with cattle production systems through a systematic review and meta-analysis of peer-reviewed publications between 2000 and 2017.

To the best of our knowledge, this is the first estimate of the overall Salmonella prevalence and the diversity of serotypes in apparently healthy cattle. We used a systematic method to identify articles reporting the prevalence of Salmonella and the serotypes in such cattle, followed by a quantitative meta-analysis to estimate the overall prevalence of Salmonella at the global level.

Salmonella colonizes the gastrointestinal tract of food animals (7) and is shed via feces (36–39). Cattle are asymptomatic carriers or reservoirs for Salmonella and may function as a source of foodborne infection (8, 23, 24). A number of serotypes frequently isolated from humans have been isolated from sick or asymptomatic cattle and some human cases have also been linked to direct exposure to cattle (20). Knowing the prevalence and diversity of Salmonella serotypes in cattle can provide important information necessary to develop preventive measures and strategies at different stages of the food chain such as application hazard analysis and critical control point (HACCP) programs in beef and milk production industries to ensure food safety (40).

There was high heterogeneity in the estimated Salmonella prevalence among the studies included in the analysis. The Salmonella prevalence can vary depending on the detection method used, the amount of sample processed, production type, and geographical variation in the distribution of the Salmonella (32, 41). The overall pooled prevalence of 9% is higher compared to other reported national level prevalence values ranging from 0.2 to 7.1% (42–46). This is not surprising since our meta-analysis provides a precise estimate (with narrow confidence interval) as it includes a higher amount of samples and total number of positive cattle for Salmonella by pooling 75 datasets from 71 publications.

The prevalence was higher in the adult cattle [9% (95%CI: 7–12%)] than in the young age group [6%(95%CI: 2–11%)]. Although the effect of age needs further investigation, this variation can presumably be in part due to variation in the number of studies included in the meta-analysis in each age group. In the young age group there were 12 publications representing only 14.2% (n = 7,477) of total cattle examined compared to 63 publications in the adult cattle with 86% of the total cattle examined. Over 70% of the publications were conducted at processing plants and in culled dairy cows destined for slaughter perhaps because of the higher public health significance at the final stage of production chain that is close to consumers (47). Even though Salmonella colonizes the intestinal tracts of cattle, there is no difference in the colonization and shedding of Salmonella between healthy calves and adult cattle (7). However, a higher prevalence of Salmonella shedding animals occurs when asymptomatic chronically infected carrier cattle are present on the farm and stay on the farm for long periods (45), which may contribute to transmission and persistence of Salmonella on the farm.

Although not statistically significant, the prevalence was higher in beef cattle compared to dairy cattle. This apparent difference can be attributed to how the animals were sampled. In most of the studies culled dairy cows were sampled at farms before shipment as opposed to beef cattle which were commonly sampled at the processing plants. Temporary restriction or complete feed withdrawal (48) and exposure to stress such as transport (42, 49, 50) can result in increased fecal shedding of Salmonella in feedlot cattle prior to slaughter.

Variations in prevalence that ranged from 2% (Europe) to 16% (North America) in various continents of the world could partly be attributed to the differences in the number of publications and the number of cattle samples included in the analysis. For North America, 26 publications (28 data sets) were retrieved consisting of 33,577 cattle samples, being the majority of the articles. In contrast, the very low prevalence estimate (2%) observed in Europe, was estimated only from 8 publications (9 data sets) in which 6,470 cattle sampled. The prevalence in South America was 14%, however this does not represent the pooled estimate as only one article was included in the analysis. The differences might also be associated with the differences in the monitoring and surveillance mechanisms among the continents (51).

Difference in the prevalence was also observed among categories of detection methods. In the majority (91%) of the studies, Salmonella was detected using traditional culturing methods which are in general considered less sensitive methods. Limited number of studies used immunomagnetic separation beads or PCR. Moreover, variation in the sensitivity of culture detection methods can influence the prevalence and consequently the observed heterogeneity (52).

In this systematic review, S. Montevideo and S. Typhimurium were the two most frequent and dominant serotypes reported where S. Montevideo was majorly reported from North America. Salmonella Typhimurium is one of the major serotypes that accounted for human clinical cases globally (10). Human infections and outbreaks due to S. Montevideo is also increasing around the globe (53)and reported in the USA, Europe, Australia, and Asia (54–56). There were differences in the most commonly reported serotypes and their proportions among different continents. Salmonella Typhimurium which is historically associated with cattle ranked number one in Africa, Asia, and Australasia. In North America and Europe, however, S. Montevideo and S. Dublin ranked number one, respectively. The implication of the shift in serotype with respect to public health requires further study. Interestingly, among the top 10 Salmonella serotypes identified in this study, S. Montevideo, S. Typhimurium, S. Anatum, S. Mbandaka, and S. Newport are among the World Health Organization's top 20 serotypes associated with human salmonellosis across the world (52). Spatial and temporal effects on the distribution and diversity of Salmonella have been reported (57, 58), which may explain the observed differences in the serotype diversity among the studies reporting Salmonella. Some of the serotypes reported in the present review were identified as the dominant serotypes elsewhere in cattle at varying proportions. For instance, in the USA, S. Newport (48.7%) and S. Typhimurium (7.1%) (59); in Ethiopia, S. Typhimurium (17.4%), S. Newport (13%) and S. Anatum (5.8%) (42), and in Europe, S. Typhimurium (38.6%) were reported to be the most frequent and dominant serotypes (60). On the contrary, none of these serotypes were reported from the national survey of Salmonella serotypes in cattle carried out in Japan (41).

All non-typhoidal Salmonella serotypes except a few serotypes which are host-specific, can potentially cause disease in humans and reside in one or more animal species (61). Salmonella serotypes were reported to be linked to several outbreaks following the consumption of contaminated beef, milk, and products thereof (62). S. Enteritidis and S. Typhimurium are the two most important serotypes transmitted from animals to humans in most parts of the world (51, 60, 63, 64). In the USA, 29 cases of diarrheal illness caused by S. Typhimurium were associated with the consumption of raw milk or raw-milk products from dairy cattle (65). During the period 1973–2011, of the 1,965 Salmonella outbreaks where a food vehicle was implicated, 96 were attributed to beef, accounting for 3,684 illnesses in USA. S. Newport and S. Typhimurium accounted for 18 and 17% of illnesses, and 29 and 18% of hospitalizations, respectively (19). The multidrug-resistant S. Typhimurium DT104 has also been associated with outbreaks related to beef contamination and resulted in hospitalization rates twice as that of other foodborne salmonellosis cases (65). From a total of 1,168 foodborne outbreaks of human salmonellosis in 2013 reported by the European member states, 1.6% of the cases were attributed to beef and beef products (60). This systematic review showed that S. Typhimurium was the most frequently reported serotype from cattle in Africa, Asia, and Australasia. Cattle could also contribute to the invasive non-typhoidal Salmonella disease in people who have contact with cattle feces. This is particularly important in regions like Africa where invasive non-typhoidal Salmonella infections are endemic as reviewed by Marks et al. (66). All the above evidence supports the importance of cattle and cattle associated serotypes for human salmonellosis.

Besides the datasets from the publications included in this review and meta-analysis, other relevant information was available in new articles that were published in the years 2017 and 2018 while the manuscript was under preparation by the authors. During this period, 6 full articles and three published abstracts representing 11 datasets were retrieved using the search engines (67–75). The majority of these studies were reported from Africa (67, 68, 70–74) except for two studies from Europe (69) and South America (75). Among the total of 5,868 cattle examined, 9.2% (554/6018), which is nearly equal to the pooled prevalence estimate, were reported to be positive for Salmonella species with different serotypes. The global level pooled prevalence of Salmonella in cattle was higher (9%) as compared to the pooled prevalence estimates of E. coli O157 (5.68%), which is also excreted by cattle showing the relative public health importance of Salmonella (76).

This study based on systematic reviews and meta-analysis provides an overall prevalence of Salmonella and serotype diversity in apparently healthy cattle at a global level. The results indicated variations in the level of Salmonella carriage in cattle across the world, and the presence of a diverse number of Salmonella serotypes. The estimated Salmonella prevalence was higher in North America. The predominant detection method is traditional culturing. Because of the possibility of Salmonella contamination of carcasses during slaughter and milk during milking, cattle can be a potential source of Salmonella and can lead to public health risk and economic loss if the necessary hygienic measures are not properly followed.

FG, GA, RA, LD, and LDZ designed the study and identified the search engines and key words for literature search. FG, GA, and LD analyzed the data. FG wrote the manuscript. GA, RA, LD, LDZ, and SG revised the manuscript.