Announcements
Expression of concern regarding paper by Park et al, published on 25 June 2015: “Epidemiological investigation of MERS-CoV spread in a single hospital in South Korea, May to June 2015”, Euro Surveill. 2015;20(25):pii=21169. It has been brought to our attention that some of the authors may not have been informed about the content of the above paper. There is a lack of clarity regarding rights to use the data. The editorial team are investigating what action needs to be taken.

Follow Eurosurveillance on Twitter: @Eurosurveillanc

Read our articles on Zika virus infection

Read our articles on mcr-1-mediated colistin resistance


In this issue


Home Eurosurveillance Edition  2015: Volume 20/ Issue 41 Article 2
Back to Table of Contents
Previous
Next

Eurosurveillance, Volume 20, Issue 41, 15 October 2015
Research article
Emborg, Teunis, Simonsen, Krogfelt, Jørgensen, Takkinen, and Mølbak: Was the increase in culture-confirmed Campylobacter infections in Denmark during the 1990s a surveillance artefact?

+ Author affiliations


Citation style for this article: Emborg HD, Teunis P, Simonsen J, Krogfelt KA, Jørgensen CS, Takkinen J, Mølbak K. Was the increase in culture-confirmed Campylobacter infections in Denmark during the 1990s a surveillance artefact?. Euro Surveill. 2015;20(41):pii=30041. DOI: http://dx.doi.org/10.2807/1560-7917.ES.2015.20.41.30041

Received:27 November 2014; Accepted:21 September 2015


Introduction

Following the description of a simple method for Campylobacter spp. isolation in 1977 [1], this genus was recognised as the leading cause of bacterial gastroenteritis in most industrialised countries [2]. The incidence of reported cases of Campylobacter infection rose dramatically during the 1990s in many industrialised countries and the reasons for this increase have been discussed in the scientific literature [3,4]. As contaminated poultry, in particular chickens, is expected to be a principal source of Campylobacter infections, it has been suggested that the emergence of these infections in a number of industrialised countries was caused by increased consumption of fresh chicken in the 1990s [5]. This hypothesis has been supported by analysis of a ‘natural experiment’, in which withdrawal of chicken and eggs from the Belgium market in June 1999 due to dioxin-contaminated feed resulted in a 40% decrease in human Campylobacter infections, as well as by interventions in New Zealand and Iceland in 1997–2008 and 1995–2007, respectively to reduce contamination of chickens [6-8]. However, it is also likely that increased diagnostic activity after the implementation of routine methods in clinical microbiology to detect Campylobacter and increased awareness by clinicians contributed to the increase in the number of Campylobacter infections [4].

In Denmark, the annual number of reported Campylobacter infections increased during the 1990s and from 1993 to 1999 a 2.5-fold increase in the incidence of Campylobacter infection was observed among individuals aged 30 years and above (Figure 1). The aim of our study was to use historical Danish serum collections of individuals representing the adult Danish population to determine the seroincidence of Campylobacter infections in 1991, 1999 and 2006. The main hypothesis was that the increase in the number of reported Campylobacter infections in Denmark reflects a real increase in incidence, and consequently the seroincidence was expected to be lower in 1991 than in 1999 as well as 2006.

Figure 1

Incidence of laboratory-confirmed Campylobacter infections per 100,000 inhabitants aged 30 years and above, Denmark, 1993–2011

/images/dynamic/articles/21277/14-00752-f1

Source: Danish Register of Enteric Pathogens.

Methods

Study population

In 1991, 1999 and 2006, three consecutive cross-sectional studies were organised by the same group of investigators [9]. Danish adults (with Danish citizenship and born in Denmark) living in one of the 11 municipalities of Copenhagen County were invited to participate in a general health examination. The study populations were randomly selected from the Danish Central Personal Register among individuals above 18 years of age.

Blood samples were taken during the following time periods: (i) February 1991 to May 1992, from 2,017 individuals; (ii) March 1999 to January 2001, from 3,501 individuals; and (iii) June 2006 to June 2008, from 3,471 individuals. These three cross-sectional studies are referred to hereafter as the 1991, 1999 and 2006 studies, respectively.

For our study presented here, 500 serum samples were randomly selected using a random-digits algorithm from each cross sectional study. In each of the three studies, 125 individuals from each of the age groups 30–39, 40–49, 50–59 and 60–69 years were included.

Serology

All the serum samples (3 × 500) were tested for IgG, IgM and IgA antibodies against Campylobacter at the serological laboratory at SSI using a direct in-house-developed Campylobacter ELISA based on a combination of C. jejuni O:1,44 and O:53 antigens in the ratio 1:1 [10].

Statistical analyses

The distribution of IgA, IgG and IgM optical density (OD) values in samples from the three studies were compared overall and in the four age groups and by sex using the Kruskal–Wallis test. If this test showed significant differences at 5% level, pairwise comparison was performed using Wilcoxon two-sample test. The statistical programme SAS version 9.4 was used for the descriptive analyses (SAS Institute, Cary, NC).

Seroincidence estimation

We have previously developed and described a Bayesian mathematical back-calculation model to estimate the incidence rate of Campylobacter seroconversion in humans [11]. This model was based on the IgG, IgM and IgA kinetics observed during a longitudinal study of 210 patients with stool culture-confirmed Campylobacter infections in Denmark from 1996 to 1997 [10]. The model incorporates inter-individual variation of peak antibody response and decay rates. By combining the information from the longitudinal study with measurements from the cross-sectional studies analysed here, we obtained estimates of the annual Campylobacter seroincidence in the population. These seroincidences were compared pairwise by calculating means and percentiles of the posterior distributions of incidence rate ratios [12].

Based on the mathematical model, a seroincidence calculator tool was developed in the statistical programme R [11] and since March 2015, this tool has been freely available from the European Centre for Disease Prevention and Control (ECDC) [13]. The tool and the underlying model to estimate seroincidences depend on a number of factors, including the level of censoring chosen for the analysis. When we use censoring, we assume that following a campylobacter infection, antibody levels do not decay towards zero but remain elevated above zero, compatible with a baseline Campylobacter antibody level a long time after the infection. These baseline antibody levels are reached after 4.5, 2.0 and 2.5 months for IgG, IgM and IgA, respectively. In the tool, censoring means that OD values below the chosen censoring level are not used to calculate the exact time since last infection, but contribute as censored observations where a long time since last infection has occurred. In the study presented here, we applied different censoring levels to illustrate the impact of censoring. In the first analysis, all three antibodies were censored at the OD value 0.25, which is a low censoring level, in particular for IgG and IgM. In the second analysis, IgG, IgM and IgA OD values were censored at 1.0, 0.4 and 0.2, respectively. This set of censoring values was chosen because a clear decrease in OD values following the acute phase of Campylobacter infection was still observed in the longitudinal data at these censoring levels [10].

Sources of supporting data

Campylobacter laboratory diagnoses are notifiable in Denmark and all culture-confirmed human cases are entered into the Danish Register of Enteric Pathogens. The number of notified cases per year in individuals aged 30 years and above was extracted from this register and the incidence per 100,000 inhabitants is shown in Figure 1.

The consumption of poultry meat in Denmark per year was obtained from Statistics Denmark [14].

Results

The proportions of men in the 500 individuals selected were 47% (n = 236), 52% (n = 259) and 45% (n = 224) in the 1991, 1999 and 2006 studies, respectively, which were not significantly different from the general population in 1991 and 1999, but the proportion of men was significantly lower in the 2006 study compared with the general population, in which 49.5% (2,685,846/5,427,459; p = 0.036) were men [15].

IgA OD values differed between the three studies (p = 0.039, Kruskal–Wallis test) with a median OD value of 0.48 (interquartile range (IQR): 0.28) in 1999 compared with 0.43 (IQR: 0.25) in 1991 and 0.42 (IQR: 0.29) in 2006. Pairwise comparisons showed significantly higher IgA OD values in 1999 compared with those in 1991 (p = 0.016) and 2006 (p = 0.0486). For IgG, a clear difference was observed between the studies (p = 0.0002, Kruskal-Wallis test), with the highest median OD value of 0.66 (IQR: 0.59) in 2006 compared with 0.58 (IQR: 0.40) in 1999 and 0.57 (IQR: 0.43) in 1991. Pairwise comparisons showed significantly higher IgG OD values in 2006 compared with those in 1991 (p < 0.001) and 1999 (p < 0.001). For IgM, no significant difference in measured OD values was observed between the years (p = 0.416, Kruskal–Wallis test) (Figure 2).

Figure 2

Measured Campylobacter IgG, IgM and IgA optical density values in cross-sectional serum samples in the 1991, 1999 and 2006 studiesa, Denmark (n = 500 per study)

/images/dynamic/articles/21277/14-00752-f2

OD: optical density.

The measured OD values are presented on a logarithmic scale. The box shows the interquartile range (IQR), the horizontal line across the box is the median, while the diamond is the mean. Dots above the upper bar represents values that are at least 1.5 times the IQR above the 75th percentile; dots below the lower bar represents values that are at least 1.5 times the IQR below the 25th percentile. Where no such dots are shown, the upper and lower bars represents the maximum and minimum values measured.

a Blood samples were taken during the following time periods: (i) February 1991 to May 1992; (ii) March 1999 to January 2001; and (iii) June 2006 to June 2008. These three cross-sectional studies are referred to as the 1991, 1999 and 2006 studies, respectively.

The distribution of measured campylobacter OD values across the studies and between age groups is shown in Figure 3. IgA OD values did not differ significantly between age groups (p = 0.487). Kruskal–Wallis test shows significant differences in IgG OD values between the age groups (p < 0.001), with the lowest median OD value of 0.54 (IQR: 0.44) in those aged 30–39 years and the highest median OD value of 0.66 (IQR: 0.52) observed among those aged 60–69 years. Pairwise comparisons showed significant differences at 5% level between all age groups except between IgG OD values from those aged 30–39 and 40–49 years, as well between the age groups 40–49 and 50–59 years. For IgM OD values, a clear age difference was observed (p < 0.001), with the highest median OD value of 0.74 (IQR: 0.50) observed among individuals in their 30s compared with those in their 60s, where the median OD value was 0.52 (IQR: 0.41). Pairwise comparisons showed significant differences at 5% level between all age groups except between IgM OD values from those aged 40– 49 and 50–59 years.

Figure 3

Measured Campylobacter IgG, IgM and IgA optical density values in cross-sectional serum samples, by age group, in the 1991, 1999 and 2006 studiesa, Denmark (n = 375 per age group)

/images/dynamic/articles/21277/14-00752-f3

OD: optical density.

The measured OD values are presented on a logarithmic scale. The box shows the interquartile range (IQR), the horizontal line across the box is the median, while the diamond is the mean. Dots above the upper bar represents values that are at least 1.5 times the IQR above the 75th percentile; dots below the lower bar represents values that are at least 1.5 times the IQR below the 25th percentile. Where no such dots are shown, the upper and lower bars represents the maximum and minimum values measured.

a Blood samples were taken during the following time periods: (i) February 1991 to May 1992; (ii) March 1999 to January 2001,; and (iii) June 2006 to June 2008. These three cross-sectional studies are referred to as the 1991, 1999 and 2006 studies, respectively.

Comparing antibody levels by sex in each of the three cross-sectional studies showed that IgG OD values were not significantly different. IgM OD values were significantly higher in women in all three studies while IgA OD values were higher in men in all three studies, but this difference was only significant in the 1991 study.

The estimated Campylobacter seroincidences in 1991, 1999 and 2006 using the two different censoring levels described above are shown in Tables 1 and 2. The seroincidence estimates changed when the censoring level changed, resulting in lower estimated seroincidences when censoring levels increased. The annual risk of at least one infection per year per person was estimated to be around 70% when the low censoring level was used and about 50% with the high level. The two different censoring levels produced the same seroincidence patterns over time and both approaches showed no significant differences in seroincidence rate between the 1991, 1999 and 2006 studies. For comparison, the incidence of laboratory-confirmed cases of Campylobacter infection in individuals aged 30 years and above from 1993 to 2011 are presented in Figure 1, showing a clear increase in the number of laboratory-confirmed cases of Campylobacter infection from 1993 to 2001.

Table 1

Estimated Campylobacter seroincidence in 1991, 1999 and 2006 studies in Denmark, and pairwise comparisons of the estimated incidence using low censoring levelsa


Study Sampling period Estimated seroincidence
per person-year
Risk of at least one infection per year per person Pairwise comparison
Incidence rate-ratios (PI)
Mean (95% PI) 1 − e−seroincidence × 1 1991–1992 1999–2001
1991 Feb 1991–May 1992 1.212 (1.131–1.299) 70%
1999 Mar 1999–Jan 2001 1.170 (1.092–1.253)  69% 1.04 (0.90–1.19)
2006 Jun 2006–Jun 2008 1.203 (1.121–1.291)  70% 1.01 (0.88–1.16) 0.97 (0.85–1.12)

PI: prediction interval.

a All three antibodies were censored at the optical density value 0.25.

Table 2

Estimated Campylobacter seroincidence in 1991, 1999 and 2006 studies in Denmark, and pairwise comparisons of the estimated incidence using high censoring levelsa


Study Sampling period Estimated seroincidence
per person-year
Risk of at least one infection per year per person Pairwise comparison
Incidence rate ratios (PI)
Mean (95% PI) 1 − e−seroincidence × 1 1991–1992 1999–2001
1991 Feb 1991–May 1992 0.696 (0.647–0.748)  50%
1999 Mar 1999–Jan 2001 0.672 (0.625–0.722)  49% 1.04 (0.90–1.20)
2006 Jun 2006–Jun 2008 0.737 (0.685–0.792)  52% 0.94 (0.82–1.09) 0.91 (0.79–1.05)

PI: prediction Interval.

a IgG, IgM and IgA optical density values were censored at 1.0, 0.4 and 0.2, respectively.

Discussion

To the best of our knowledge, this is the first study in which serology has been used to estimate Campylobacter seroincidence over time and in which this measure has been compared with the number of notified cases during the same period. Although there were minor differences in IgA and IgG values between samples from the three studies, these did not result in significant differences between the seroincidence rates. This is in contrast to the reported number of laboratory-confirmed cases of Campylobacter infection, for which a 2.5-fold increase in incidence per 100,000 inhabitants aged 30 years or more was observed, from 21.5/100,000 in 1993 (677 cases) to 54.8/100,000 in 1999 (1,821 cases) (Figure 1). Thuesen et al. [16] looked further into who participated in the 2006 study and they found they were older, had a higher educational level and higher income, while non-responders were often living alone, were men, had a higher prevalence of hospitalisation and more days at hospital for any reason. On the other hand, use of prescription drugs and the prevalence of more than one annual contact with general practitioners were higher among responders. Bender et al. [17] found that participants in the 1999 study cohort were older, were more often house owners, wage earners, living with a partner, had a higher education level and higher income. These studies indicate that people who participate in general health examinations have a higher socioeconomic status and they might also be individuals with an interest in a healthy lifestyle. A similar study was not carried out for the 1991 study cohort; however, there is no strong indication that the effect of selection bias will have changed over the years covered in the studies, indicating that the responders were comparable over time.

The seroincidence rates presented in Tables 1 and 2 are very high compared with the number of laboratory-confirmed cases of Campylobacter infection (Figure 1). Even when the high levels of censoring were used, the risk of at least one infection per year per person was about 50%, which suggests that the immune system of the study participants aged 30 years and above had been exposed to Campylobacter every other year. The seroincidence does not reflect number of clinically ill individuals and it is conceivable that only a fraction of Campylobacter infections will lead to clinical illness [18]. Given the fact that Campylobacter is ubiquitous in many environmental reservoirs as well as in poultry, it is biologically plausible that the seroincidence is much higher than the number of reported cases, as has been observed for Salmonella infections [12,19-21].

In principle, the model may be able to assign a time since infection (and a corresponding distribution) for all cases of Campylobacter infection, even when measured antibody levels are low. However, the back-calculation is more precise in a short time period after infection, i.e. within a couple of months, when antibody levels are generally high, rather than several months or years after presumed infection. On this basis, censoring at high OD values that correspond to a short estimated time since infection may be preferable, due to uncertainties in the seroincidence measurement. Other researchers, however, may prefer to include all available data and present seroincidence rates accordingly, but with larger prediction intervals.

Seroresponse following Campylobacter infection may indicate some degree of immunity against infection. In two studies from the United States, students developed gastrointestinal illness following the consumption of unpasteurised milk [22,23]. These outbreaks were caused by C. jejuni, and ill students had antibody levels consistent with recent infection. In both studies, there were groups of individuals who had high C. jejuni antibody levels who did not become ill. These high antibody levels without illness were correlated with habitual consumption of unpasteurised milk, which indicates the development of immunity [22,23]. In an experimental study of C. jejuni infection in humans in the United States, reinfection of volunteers with the same strain did not result in clinical illness, supporting the theory that at least short-term immunity is developed following infection [18]. This underscores that seroincidence rates are not a direct measure of clinical illness but rather a measure of the force of infection at the population level, and that this, to some extent, may be related to immunity.

Studies from Thailand in 1980 and 1987 showed that repeated Campylobacter infections are common in early childhood (<24 months-old) and infection rates decrease with age, paralleled by a progressive increase in specific serum antibodies [24,25]. Studies from industrialised countries also indicate the importance of age. A Danish study of the spatial distribution of Campylobacter infections found that residence in rural areas and areas with a low population density were both associated with an increased risk of infection, and that this association concerned children aged 0–14 years in particular. This association could explain a third of cases among children in the Danish countryside in 1991 to 2001 [26]. Furthermore, in Wisconsin in the United States, a seroepidemiological study in 1997 to 1999 among rural children showed that increasing age as well as farm residence were associated with increasing C. jejuni seropositivity. In the age group 15–18 years, between 85% and 90% of farm-resident children were seropositive, while for children not living on farms, it was between 60% and 65% [27]. Also, in the Netherlands, a high level of Campylobacter antibodies and a high seroincidence has been measured from 30 years of age [28]. In the study presented here, all participants were aged 30 years and above and only small differences in measured antibody OD values were observed between the age groups investigated. This corroborates the hypothesis that Danish children and young adults are exposed to Campylobacter, after which specific antibodies are produced and the antibody level remains elevated. The age dependency of the IgM response, which reappears less strongly after repeated infections, supports this notion.

In addition to poultry and cattle, major Campylobacter transmission routes include animal contact with farm animals, occupations related to farm animals and contact with environment connected to farms [29]. With many different reservoirs harbouring Campylobacter and a seasonal variation in the number of Campylobacter-infected broiler flocks, with up to 80% of flocks infected during summer, it is likely that many Danes are exposed at regular intervals to Campylobacter, followed by seroconversion, although they do not become clinically ill [30,31].

Our group has previously shown that the emergence of Salmonella Enteritidis in the 1990s was mirrored by an almost parallel increase in the seroincidence of Salmonella [21]. We expected to see a similar pattern for Campylobacter, which was considered as an emerging infection in the late 1990s. To our surprise, however, we were unable to demonstrate an increase in the Campylobacter seroincidence rates, which raises the hypothesis that at least part of the increased reporting of Campylobacter infections may have been due to increased and improved diagnostic activity and awareness, rather than a true increase in incidence. There are, however, some important counter arguments. First of all, the increase in the number of reported culture-confirmed cases during the 1990s in Denmark was similar to the increases seen in other European countries, and these increases have been linked to increased consumption of fresh poultry meat [3,4]. A Danish case–control study from 2000 to 2001 found that the main domestic risk factor for campylobacteriosis was consumption of chicken meat that had been bought fresh and subsequently not frozen [5]. In 2011, source attribution of Campylobacter in Denmark indicated that Danish chicken and cattle as well as imported chicken are important sources of infection [29]. The association between poultry and Campylobacter infections in humans is corroborated by the fact that during the 1990s, the consumption of poultry meat in Denmark increased from 12.4 kg per resident in 1991 to 19.2 kg in 2000 [14]. In addition, the consumption of fresh poultry meat increased during the same period [32]. As freezing kills Campylobacter bacteria, the change in consumer habits towards more poultry meat per inhabitant per year and consumption of fresh poultry meat has most likely increased the Campylobacter dose that humans are exposed to [5]. In addition, the proportion of Danish broiler flocks testing positive for Campylobacter is highest in August and lowest during winter [30,31], which is the same seasonal pattern as human Campylobacter infections [31]. A previous study has shown that the quantity of Campylobacter-contaminated food products consumed was directly related to the occurrence and severity of disease [22]. This indicates that the increase in consumption of fresh poultry during the 1990s was a possible explanation why the reported number of human cases of Campylobacter infection started to increase without any change in seroincidence rate during the same period, as Campylobacter antibody levels were already at a high level. Furthermore, the proportion of reported cases is very small compared with the estimated seroincidence. We propose that the increase in rates of clinical illness was driven by the increased consumption of fresh poultry, whereas seroincidence is a composite measure reflecting various sources of exposures, of which poultry is merely a fraction. As an example, a number of large outbreaks of Campylobacter were recently attributed to contaminated drinking water [33,34].

Use of serology as a tool to measure Campylobacter infections in the population is also subject to limitations. The longitudinal study described IgG, IgA and IgM Campylobacter antibody decay profiles of 210 patients with stool culture-confirmed campylobacter infections and it is assumed that the seroresponse in the longitudinal study population is similar to that in the cross-sectional populations [11]. In addition, the seroincidence calculator tool can only be used when the cross–sectional samples are analysed using the same ELISA method as that used to analyse the sera from the longitudinal study. Finally, we cannot rule out that we might have observed different seroincidences in the 1991, 1999 and 2006 studies if children and adolescents had been chosen as study populations.

In conclusion, we were unable to associate the emergence of culture-confirmed Campylobacter infections with a similar increase in seroincidence at the population level. This suggests that Campylobacter was widely present in the Danish population before the increased incidence of poultry-associated clinical human Campylobacter infections and that other sources of infection were present. Given our findings, there may be a larger diversity of Campylobacter exposure than hitherto thought; however, the development of clinical illness might be related to the Campylobacter dose individuals are exposed to.


Acknowledgements

We would like to thank the European Centre for Disease Prevention and Control (ECDC) that funded this work. We acknowledge Andrea Ammon from ECDC for her constant support to the ECDC-funded seroepidemiology project. We also acknowledge Bjørn Kantsø and Daniel Silla Jatta from the Division of Microbiological Diagnostics and Infection Control at Statens Serum Institut for their efficient laboratory work.

Conflict of interest

None declared.

Authors’ contributions

Hanne-Dorthe Emborg led the writing of the paper and was responsible for interpretation of the obtained results. All authors provided contributions to the paper and approved the final version. Peter Teunis developed the seroincidence calculator tool used to estimate Campylobacter seroincidences. Jacob Simonsen performed the statistical analyses. Kåre Mølbak got the original idea for this study. Karen A. Krogfelt and Charlotte S. Jørgensen were responsible for the laboratory work and provided input to the interpretation of results. Johanna Takkinen provided input to the discussion and interpretations of the obtained results.


References

  1. Skirrow MB. Campylobacter enteritis: a “new” disease.BMJ. 1977;2(6078):9-11.

  2. European Food Safety Authority (EFSA), European Centre for Disease Prevention and Control (ECDC). The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2010. EFSA Journal. 2012;10(3):2597. Available from: http://www.efsa.europa.eu/sites/default/files/scientific_output/files/main_documents/2597.pdf

  3. Allos BM. Campylobacter jejuni Infections: update on emerging issues and trends.Clin Infect Dis. 2001;32(8):1201-6.

  4. Olson CK, Ethelberg S, Van Pelt W, Tauxe RV. Epidemiology of Campylobacter jejuni infections in industrialized nations. In: Nachamkin I, Szymanski CM, Blaser MJ, editors. Campylobacter. 3rd ed. Washington DC: ASM press; 2008. p. 163-89.

  5. Wingstrand A, Neimann J, Engberg J, Nielsen EM, Gerner-Smidt P, Wegener HC,  et al.  Fresh chicken as main risk factor for campylobacteriosis, Denmark. Emerg Infect Dis. 2006;12(2):280-5.DOI: 10.3201/eid1202.050936 PMID: 16494755

  6. Vellinga A, Van Loock F. The dioxin crisis as experiment to determine poultry-related campylobacter enteritis.Emerg Infect Dis. 2002;8(1):19-22.DOI: 10.3201/eid0801.010129 PMID: 11749743

  7. Tustin J, Laberge K, Michel P, Reiersen J, Dađadóttir S, Briem H,  et al.  A national epidemic of campylobacteriosis in Iceland, lessons learned. Zoonoses Public Health. 2011;58(6):440-7.DOI: 10.1111/j.1863-2378.2010.01387.x PMID: 21824341

  8. Sears A, Baker MG, Wilson N, Marshall J, Muellner P, Campbell DM,  et al.  Marked campylobacteriosis decline after interventions aimed at poultry, New Zealand. Emerg Infect Dis. 2011;17(6):1007-15.DOI: 10.3201/eid/1706.101272 PMID: 21749761

  9. Thyssen JP, Linneberg A, Menné T, Nielsen NH, Johansen JD. The prevalence and morbidity of sensitization to fragrance mix I in the general population.Br J Dermatol. 2009;161(1):95-101.DOI: 10.1111/j.1365-2133.2009.09157.x PMID: 19438476

  10. Strid MA, Engberg J, Larsen LB, Begtrup K, Mølbak K, Krogfelt KA. Antibody responses to Campylobacter infections determined by an enzyme-linked immunosorbent assay: 2-year follow-up study of 210 patients. Clin Diagn Lab Immunol.2001;8(2):314-9.

  11. Teunis PFM, van Eijkeren JCH, Ang CW, van Duynhoven YTHP, Simonsen JB, Strid MA,  et al.  Biomarker dynamics: estimating infection rates from serological data. Stat Med. 2012;31(20):2240-8.

  12. Simonsen J, Teunis P, van Pelt W, van Duynhoven Y, Krogfelt KA, Sadkowska-Todys M,  et al.  Usefulness of seroconversion rates for comparing infection pressures between countries. Epidemiol Infect. 2011;139(4):636-43.DOI: 10.1017/S0950268810000750 PMID: 20380770

  13. Teunis P. Seroincidence: estimating infection rates from serological data. R package version 1.0.0. Stockholm: European Centre for Disease Prevention and Control; 2015. Available from: http://ecdc.europa.eu/en/data-tools/seroincidence-calculator-tool/Pages/default.aspx

  14. Statistics Denmark. Prices and consumption, Food consumption. 2015. Copenhagen: Statistics Denmark. [Accessed 24 Aug 2015]. Available from: http://www.statbank.dk/statbank5a/default.asp?w=1920

  15. Statistics Denmark. Population and Elections, Population and population projections, Population in Denmark. 2015. Copenhagen: Statistics Denmark. [Accessed 24 Aug 2015]. Available from: http://www.statbank.dk/statbank5a/default.asp?w=1920

  16. Thuesen BH, Cerqueira C, Aadahl M, Ebstrup JF, Toft U, Thyssen JP,  et al.  Cohort Profile: the Health2006 cohort, Research Centre for Prevention and Health. Int J Epidemiol. Int J Epidemiol. 2014 Apr;43(2):568-75.PMID: 23615486

  17. Bender AM, Jørgensen T, Helbech B, Linneberg A, Pisinger C. Socioeconomic position and participation in baseline and follow-up visits: the Inter99 study.Eur J Prev Cardiol. 2014;21(7):899-905.DOI: 10.1177/2047487312472076 PMID: 23233551

  18. Black RE, Levine MM, Clements ML, Hughes TP, Blaser MJ. Experimental Campylobacter jejuni infection in humans.J Infect Dis. 1988;157(3):472-9.

  19. Falkenhorst G, Simonsen J, Ceper TH, van Pelt W, de Valk H, Sadkowska-Todys M,  et al.  Serological cross-sectional studies on salmonella incidence in eight European countries: no correlation with incidence of reported cases. BMC Public Health. 2012;12(1):523.DOI: 10.1186/1471-2458-12-523 PMID: 22799896

  20. Simonsen J, Mølbak K, Falkenhorst G, Krogfelt KA, Linneberg A, Teunis PFM. Estimation of incidences of infectious diseases based on antibody measurements.Stat Med. 2009;28(14):1882-95.

  21. Simonsen J, Strid MA, Mølbak K, Krogfelt KA, Linneberg A, Teunis P. Sero-epidemiology as a tool to study the incidence of Salmonella infections in humans.Epidemiol Infect. 2008;136(7):895-902.DOI: 10.1017/S0950268807009314 PMID: 17678562

  22. Blaser MJ, Sazie E, Williams LP. The influence of immunity on raw milk--associated Campylobacter infection.JAMA. 1987;257(1):43-6.

  23. Jones DM, Robinson DA, Eldridge J. Serological studies in two outbreaks of Campylobacter jejuni infection.J Hyg (Lond). 1981;87(2):163-70.DOI: 10.1017/S0022172400069369 PMID: 6895231

  24. Blaser MJ, Taylor DN, Echeverria P. Immune response to Campylobacter jejuni in a rural community in Thailand.J Infect Dis. 1986;153(2):249-54.DOI: 10.1093/infdis/153.2.249 PMID: 3944480

  25. Taylor DN, Perlman DM, Echeverria PD, Lexomboon U, Blaser MJ. Campylobacter immunity and quantitative excretion rates in Thai children.J Infect Dis. 1993;168(3):754-8.DOI: 10.1093/infdis/168.3.754 PMID: 8354916

  26. Ethelberg S, Simonsen J, Gerner-Smidt P, Olsen KEP, Mølbak K. Spatial distribution and registry-based case-control analysis of Campylobacter infections in Denmark, 1991-2001.Am J Epidemiol. 2005;162(10):1008-15.

  27. Belongia EA, Chyou P-H, Greenlee RT, Perez-Perez G, Bibb WF, DeVries EO. Diarrhea incidence and farm-related risk factors for Escherichia coli O157:H7 and Campylobacter jejuni antibodies among rural children.J Infect Dis. 2003;187(9):1460-8.DOI: 10.1086/374622 PMID: 12717628

  28. Ang CW, Teunis PFM, Herbrink P, Keijser J, Van Duynhoven YHTP, Visser CE,  et al.  Seroepidemiological studies indicate frequent and repeated exposure to Campylobacter spp. during childhood. Epidemiol Infect. 2011;139(9):1361-8.DOI: 10.1017/S0950268810002359 PMID: 20974019

  29. Annual report on zoonoses in Denmark 2011. The Danish Zoonosis Centre, National Food Institute, Technical University of Denmark; 2012. Available from: http://www.food.dtu.dk/english/-/media/Institutter/Foedevareinstituttet/Publikationer/Pub-2012/Annual%20report%20on%20Zoonoses%20in%20Denmark%202011.ashx?la=da

  30. Annual report on zoonoses in Denmark 2000. Ministry of Food, Agriculture and Fisheries; 2001. Available from: http://www.food.dtu.dk/english/-/media/Institutter/Foedevareinstituttet/Publikationer/Pub-2001/annual_report_zoonoses_2000.ashx?la=da

  31. Annual report on zoonoses in Denmark 2007. The Danish Zoonosis Centre, National Food Institute, Technical University of Denmark; 2009. Available from: http://www.food.dtu.dk/english/-/media/Institutter/Foedevareinstituttet/Publikationer/Pub-2008/annual_report_zoonoses_2007.ashx?la=da

  32. Christensen B, Sommer H, Rosenquist H, Nielsen N. Risk assessment on Campylobacter Jejuni in chicken product. The Danish Veterinary and Food Administration Institute of Food Safety and Toxicology Division of Microbiological Safety; 2001. Available from: http://www.researchgate.net/publication/277283652_Risk_Assessment_on_Campylobacter_Jejuni_in_Chicken_Product

  33. Gubbels S-M, Kuhn KG, Larsson JT, Adelhardt M, Engberg J, Ingildsen P,  et al.  A waterborne outbreak with a single clone of Campylobacter jejuni in the Danish town of Køge in May 2010. Scand J Infect Dis. 2012;44(8):586-94.DOI: 10.3109/00365548.2012.655773 PMID: 22385125

  34. Braeye T, DE Schrijver K, Wollants E, van Ranst M, Verhaegen J. A large community outbreak of gastroenteritis associated with consumption of drinking water contaminated by river water, Belgium, 2010.Epidemiol Infect. 2015;143(4):711-9.



Back to Table of Contents
Previous
Next

The publisher’s policy on data collection and use of cookies.

Disclaimer: The opinions expressed by authors contributing to Eurosurveillance do not necessarily reflect the opinions of the European Centre for Disease Prevention and Control (ECDC) or the editorial team or the institutions with which the authors are affiliated. Neither ECDC nor any person acting on behalf of ECDC is responsible for the use that might be made of the information in this journal. The information provided on the Eurosurveillance site is designed to support, not replace, the relationship that exists between a patient/site visitor and his/her physician. Our website does not host any form of commercial advertisement. Except where otherwise stated, all manuscripts published after 1 January 2016 will be published under the Creative Commons Attribution (CC BY) licence. You are free to share and adapt the material, but you must give appropriate credit, provide a link to the licence, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.

Eurosurveillance [ISSN 1560-7917] - ©2007-2016. All rights reserved.
 

This website is certified by Health On the Net Foundation. Click to verify. This site complies with the HONcode standard for trustworthy health information:
verify here.