From the clinical infectious disease database, we extracted 4,269 individual reports (only 1 per person) of men who had acquired a notifiable disease through MSM contact during January 1, 2010–December 31, 2018. Of these, 83 (1.9%) men were excluded because they did not have either a valid CPR number or an address in Denmark. For the control group, we extracted 15,250 randomly selected matched male controls from the CPR registry. The mean age was 41 years (median 40 years, range 18–88 years) for both the case-patients and the controls. Case-patients and controls were geographically distributed throughout Denmark, but a significantly larger proportion of the study population than the general population resided in the Capital Region (χ² = 1,400; p<0.0001).

From the Register of Enteric Pathogens, we extracted 49,321 notifications of infections with >1 of the 3 bacterial pathogens (exposures); 748 patients had registrations for >2 recurrent infections with the same pathogen during the study period (Table 1). Most (76%) notifications were for Campylobacter infections: 55% C. jejuni, 4% C. coli, and 40% other Campylobacter spp. with no species reported. In the MSM group, a total of 132 Campylobacter, 3 Salmonella, and 64 Shigella infections were reported during the study period (Table 1). In the control group, we observed 74 Campylobacter, 44 Salmonella, and 4 Shigella infections (Table 1). Compared with controls, a higher proportion of MSM case-patients <40 years of age had Campylobacter infections (Table 1), although this difference was not statistically significant (χ² = 3; p = 0.08).

Figure 2

Figure 2. Percentages of clinical notifications of infections acquired through MSM contact (notifiable infections) and Campylobacter, Salmonella, and Shigellainfections reported among MSM and controls in study...

Overall, we found that the odds of a Campylobacter infection were 14 times higher among MSM than controls (Table 2). MSM case-patients also had 74 times higher odds than controls of being infected with Shigella, a pathogen known to be transmitted by sexual contact (Table 2). However, we found no significant difference between Salmonella infection rates in MSM case-patients and controls (Table 2). MSM case-patients who were infected with Campylobacter were significantly less likely to have acquired their infection abroad compared with controls (χ² = 21; p<0.001), which was not the case for Salmonella or Shigella. Over the study period, there were 4 (3%) recurrent Campylobacter infections in the MSM group compared with none in the control group (statistical analysis not possible); there were no recurrent Salmonella or Shigella infections in the MSM or control groups. During the study period, we observed an increase among MSM case-patients in clinical infections acquired through MSM contact and Campylobacter or Shigella infections (Figure 2, t = 5–11; p<0.001). We did not observe any change in the proportion of Salmonella among MSM case-patients (t = −2; p = 0.1) or for any of the 3 pathogens in the control group (Figure 2, t = −1 to 2; p = 0.1–0.4).

We present surveillance data–driven evidence that campylobacteriosis can be transmitted through sexual contact. Among our study group of MSM case-patients, the odds of being infected with Campylobacter was 14 times higher than among controls. We observed a similar pattern for shigellosis, which is known to be transmitted by sexual contact, but not for Salmonella infection. In Denmark, Salmonella infections are considered almost exclusively foodborne whereas domestic Shigella infections are regarded as primarily sexually transmitted and secondarily as foodborne or transmitted through general person-to-person contact. Campylobacter infection is linked to several different transmission routes (4), handling or consuming of poultry long considered the most notable.

However, our results reinforce information from many reports suggesting Campylobacter can be transmitted through sexual contact; these reports provide explanations for unique aspects of Campylobacter epidemiology supported by biologic facts from other foodborne bacteria. Reported outbreaks among MSM in Canada (14,19,23,24) and previous observations of higher infection rates in homosexual men and HIV-positive patients (8,13,17,20,21,25,26) indicate the likelihood that Campylobacter can be transmitted through sexual contact. In spite of this, sexual contact has not traditionally been considered a possible transmission route in sporadic campylobacteriosis cases and therefore has been excluded as a possible risk factor in published case-control studies (4). We were unable to assess the risk of Campylobacter transmission through anal–oral contact between heterosexual partners because only MSM sexual contact is reported on notifications. However, it is highly likely that the risk of infection through this type of sexual contact is equally relevant for heterosexual and MSM partners.

Several aspects of Campylobacter epidemiology remain to be clarified, such as why the disease is more common among men (3,13,27–29). Many explanations have been proposed, including differences in food handling and preparation, healthcare-seeking behaviors, and physiologies. We provide an additional explanation: adult men practicing sex with other men are at significantly higher risk (p<0.001) of campylobacteriosis. In this study, Campylobacter incidence in the MSM group was almost 5 times higher than in the general population, which could be a partial driver for higher incidence in men.

Another feature observed in surveillance statistics from several countries was the biphasic age distribution of Campylobacter infections, showing pronounced peaks in children <5 years of age and young adults 20–40 years of age (3,30–33). In the 20–40-year age group, these peaks have also been observed for Shigella (34,35), but not for other foodborne bacteria, such as Salmonella. These findings have been explained by the secondary weaning phase, meaning that young adults away from home are less aware of proper hygiene practices, secondary infections from caring for young children at home, or hormonal and behavioral factors. When comparing different age groups, we saw higher Campylobacter infection rates in MSM <40 years of age compared with controls, although this difference was not statistically significant because of the small sample size. On the basis of our results, we suggest that, in countries with a clear biphasic Campylobacter age distribution, some of the high infection frequency among young adults might be explained by sexual transmission among persons in this very sexually active age group.

Bacterial pathogens can be transmitted through fecal–oral contact during sex, either directly, through anal–oral contact or anal–penile–oral contact, or indirectly, such as through the use of sex toys or fingers. In 1974, sexual contact was recognized as a risk factor for infection with Shigella spp. (11), and such contact is now widely acknowledged as one of the most important transmission routes for shigellosis. The probability of infection with a foodborne bacterium from fecal–oral contact is directly related to the infectious dose. For Shigella spp. this dose can be as low as 10–1,000 organisms, and for Campylobacter, 500–10,000 organisms (36). The infectious dose for Salmonella spp. varies (37) but might be as high as 1 million bacterial cells. Pathogens with a low infectious dose are easier to transmit from person to person, including through sexual contact; on the basis of our findings, we tentatively propose that this might be the case for Shigella and Campylobacter, but not Salmonella.

In Denmark and the rest of Europe, campylobacteriosis and shigellosis are often acquired abroad; 30%–50% of infections are reported as travel related (3,33). This possibility was reflected in our control group, but Campylobacter infections among MSM were significantly less likely to have been travel related (p<0.001). Assuming that many Campylobacter and Shigella cases in the MSM group have been acquired by sexual contact, this domestic pattern is not surprising because MSM contact is most commonly not a holiday experience and infection is more likely to happen at home.

During the study period, there was a significant (p<0.001) increase in the proportion of MSM case-patients infected with Shigella or Campylobacter, but not with Salmonella; we found no increase in the control group. Incidence of campylobacteriosis in Denmark has increased recently (3), reflecting a combination of actual increases in cases of the disease, changes in diagnostic techniques, and improved electronic notification. During the study period, laboratory diagnosis of gastrointestinal pathogens has transitioned from culturing to more sensitive PCR testing of feces. This change could explain some of the increase in recorded Shigella and Campylobacter infections; however, similar increases would be expected for Salmonella and for the number of infections in the control group. An increasing trend over time would also be observed if there had been larger outbreaks of Shigella and Campylobacter among MSM case-patients later in the study period, but only 1 outbreak of shigellosis was reported, and none of campylobacteriosis among MSM during this time (38). The most likely explanation for the observed increase in Shigella and Campylobacter among MSM case-patients is the concurrent increase in the proportion of infections transmitted through MSM contact reported to SSI. More extensive reporting has increased the amount of data available in the electronic databases, in turn increasing the likelihood of discovering infections in the MSM group.

Our study had 2 major strengths: use of routinely collected high-quality national surveillance data for our analyses and results supported by biologic evidence and demographic information, such as the infectious doses and the low level of travel-related infections in the MSM group that explain some of the unique characteristics of Campylobacter transmission. The first limitation is that, although clinical notifications of diseases in Denmark can include MSM contact as possibly associated with the circumstances of the infection, such inclusions are based on patient statements, and only HIV, syphilis, gonorrhea, and hepatitis infections must be reported. Therefore, the MSM population group probably did not include all men with an infection acquired through MSM contact during the study period. It is also very likely that the control group included some MSM and that some reported Campylobacter, Shigella, or Salmonella infections in the control group were acquired through MSM contact. However, such an underestimation would tend to lower the estimates of association toward 1 and, in fact, approximate the odds ratios to risk ratios. Differences between case-patients and controls concerning how exposure was determined (i.e., from a notification to SSI about infection with an enteric pathogen) could have affected the results. However, because physicians and laboratories in Denmark operate under the same guidelines and procedures and the notification records are uniformly registered, such differences are unlikely to have had an effect.

The results in this study provide a measure of the risk associated with a notifiable infectious disease among a specific group of MSM and not among a general MSM population. However, considering the robust mORs and corresponding CIs combined with published reports about high Campylobacter incidence in MSM populations, we believe that our findings can be extrapolated to MSM in general. The case-patient group might overrepresent risk-taking persons more likely to practice unsafe sex, and if that is the case, the increased risk for Campylobacter infection through sexual contact would primarily apply to this high-risk group. However, it can also be argued that this risk applies to the wider MSM population because condom use would not eliminate the infection risk associated with several fecal–oral transmission routes. The case-patients we selected might also have a high frequency of known risk factors, such as use of proton pump inhibitors, certain occupations, or increased contact with animals, that we could not account for in this study.

Another limitation of the data used in our study was the incomplete travel information for persons with reported infections. Travel history was unknown for patients with 25% of Salmonella, 55% of Shigella, and 57% of Campylobacter infections because of incomplete registration and lack of follow-up interviews. This proportion was similar between case-patients and controls. In spite of this, our results still indicated that MSM are less likely to have acquired Campylobacter or Shigella infections abroad. Also, although our data covered all infections reported as acquired through MSM contact and all confirmed cases of campylobacteriosis, salmonellosis, and shigellosis, these do not account for all cases. Notifications are based on passive surveillance and underestimate the true incidence because persons with infections who do not seek medical attention are not captured (39). However, because our system enabled linkage of individual data collected, we believe that these results are as accurate as possible and could be generalized to similar settings. Lack of association between MSM contact and Salmonella infection might reflect the true situation or that a larger sample size is needed.

The cycle of transmission through sexual contact among adults contributes substantially to the burden of disease in highly industrialized countries, but likely not in low-middle income settings where limited hygiene and sanitation constitute greater risk factors. Therefore, the external validity of the results in some locations should be considered and we encourage public health agencies in other countries to replicate this study to corroborate the findings. Finally, as we did not directly measure Campylobacter infection transmitted through anal-oral contact, the association is speculative and our conclusions are derived from surveillance data analyses in combination with knowledge gained from studies of other bacteria.

Our findings indicate a strong likelihood that Campylobacter can be transmitted during sexual contact. Given previous reports of outbreaks and high incidence of Campylobacter among MSM, this is not surprising. Combining this theory with high-quality national surveillance data, our results offer additional reasonable explanations for why surveillance statistics from some countries show that adult men are more frequently infected with Campylobacter than are women and why Campylobacter incidence peaks among young adults. Overall, our findings not only address epidemiologic questions but also highlight a need to inform the public about the risk of infection through sexual activity in regions where Campylobacter incidence exhibits patterns similar to those in Denmark. These results are primarily applicable to adults in high-income settings in some countries; further studies are warranted among the general MSM population rather than only those with a reported notifiable disease. However, considering the high burden of campylobacteriosis in Europe, the United States, and Australia, incorporating knowledge from these findings into general information campaigns might encourage the use of precautionary measures during sexual contact, which could, in turn, lead to lower infection rates and reduced overall costs to society.