Data on all Campylobacter isolates obtained from fecal or lower gastrointestinal tract samples, reported in England and Wales from 1990 through 2007, were extracted from the national laboratory database (LabBase) and stored in a Microsoft Access (Microsoft Corp., Redmond, WA, USA) database. Cases were assigned to 10-year age groups and to geographic areas on the basis of laboratory location (northern, mid-country, southern). The season was assigned on the basis of the earliest available specimen date (spring, March–May; summer, June–August; autumn, September–November; winter, December–February). Data on cases of cryptosporidiosis and nontyphoidal salmonellosis for the same period were extracted, grouped, and categorized and manipulated as above for comparative purposes. Denominator data for England and Wales for the same period were obtained from the Office for National Statistics. Data were analyzed in Microsoft Excel 2003 and Stata version 10 (StataCorp, College Station, TX, USA). Estimates of incidence per 100,000 population were calculated throughout; relative risks (RR) and 95% confidence intervals (CIs) were calculated as required.

Figure 1

Figure 1. Incidence of laboratory-reported campylobacteriosis, England and Wales, by age group, 1990–2007.

Figure 2

Figure 2. Relative incidence of campylobacteriosis by sex, region, and season, compared with rates of salmonellosis and cryptosporidiosis, among patients >60 years of age, England and Wales, 1991–2007. Northern, northwest, northeast, as...

From 1990 through 2007 in England and Wales, 838,436 cases of Campylobacter infection were reported; patient age was available for 810,632 case-patients (96.7%). From 1990 through 1999, incidence increased in all age groups (Figure 1), but the increase was proportionate to increasing age (0–9 years of age RR 1.07 [95% CI 1.03–1.10]; 10–19 years RR 1.47 [95% CI 1.41–1.55]; 20–59 years RR 1.78 [95% CI 1.75–1.81]; >60 years RR 2.51 [95% CI 2.41–2.61]). From 2000 through 2004, incidence declined in all age groups. However, although the degree of decline was similar for those 0–9 years of age (RR 0.77 [95% CI 0.74–0.8]), 10–19 years (RR 0.73 [95% CI 0.70–0.76]), and 20–59 years (RR 0.75 [95% CI 0.74–0.76]), for patients >60 years of age, the degree of decline was significantly lower (RR 0.88 [95% CI 0.86–0.91]; χ² p<0.001). Finally, although the incidence increased only moderately among those 10–19 years of age (RR 1.02 [95% CI 0.98–1.07]) and 20–59 years (RR 1.04 [95% CI 1.03–1.06]) from 2005 through 2007, greater increases were observed for those 0–9 years (RR 1.12 [95% CI 1.08–1.17]) and those >60 years (RR 1.33 [95% CI 1.29–1.36]). During the surveillance period, therefore, the incidence among patients >60 years of age, compared with the incidence among younger patients, increased markedly (RR 0.45 [95% CI 0.44–0.47] in 1990 to RR 1.17 [95% CI 1.15–1.19] in 2007). This effect was observed for campylobacteriosis in both sexes, in 3 geographic areas of England and Wales, and in all 4 seasons, but was not observed for nontyphoidal salmonellosis or cryptosporidiosis (Figure 2; Technical Appendix).

We report a striking change in the population at risk for campylobacteriosis in England and Wales, which is independent of gender, geography, or season. The absence of a similar change in the age distribution of laboratory-reported salmonellosis or cryptosporidiosis from the same population suggests that this is unlikely to be a surveillance artifact.

Campylobacter infections are rarely typed beyond the genus level in England and Wales, so infections are routinely reported as Campylobacter species. Therefore, changes in the incidence of the various species that constitute this broad case definition could possibly explain some of the altered disease pattern reported. For example, infections caused by certain species (e.g., C. fetus) are more often associated with coexisting conditions that might occur more frequently in the elderly. This circumstance was unlikely to have affected the results of this study, however. First, isolation methods used in England and Wales favor the growth of C. jejuni and C. coli (4,5) over that of other species, including C. fetus. Furthermore, C. fetus normally causes systemic infections detected through blood culture, and the proportion of blood to fecal isolations of Campylobacter species in patients >60 years of age reported in England and Wales remained constant from 1990 through 1999 (275/58,139 fecal isolations; 0.47%), from 2000 through 2004 (185/44,349; 0.42%), and from 2005 through 2007 (135/31,637; 0.42%) (Health Protection Agency, unpub. data).

When disease incidence was ranked according to specific population group, children 0–9 years of age had the highest ranking in 1990, but by 2007, incidence for this age group ranked seventh of 9 age groups. Incidence among children <10 years declined most rapidly from 1998 to 2003. This finding led to the hypothesis that the introduction and increased utilization of the National Health Service’s NHS Direct (a 24-hour telephone, online, and interactive digital TV service, which provided health advice and information) at this time had a “triage effect” on those seeking care for children in this age group. The 2 events are correlated in time (initial NHS Direct pilot sites began taking calls in March 1998; by April 1999, 40% of the population of England had access, and by November 2000, the service was available throughout England and Wales [6]), NHS Direct has had a demonstrable negative effect on the use of general practice (7), and infants and young children are overrepresented among calls to NHS Direct about gastrointestinal conditions (8). The second Infectious Intestinal Disease study in England, currently under way (9), will provide further information upon which to assess this hypothesis, which is not readily testable by using laboratory data.

By far the most striking finding of this study is the emergence of older persons as the population most at risk for campylobacteriosis in England and Wales. Although the elderly were not the only group at risk in 2007 (because of increasing incidence), the overall trend singles them out as the main emerging at-risk group (the increase in other age groups requires continued monitoring, however). The pattern of infection in older patients is perhaps predictable, given the similar pattern for incidence of listeriosis in England and Wales since 2001 (10). As life expectancy increases in the United Kingdom, the number of persons living with chronic conditions is likely to increase; these factors suggest that the incidence of campylobacteriosis in older persons will continue to increase in the future. Therefore, risk factors for Campylobacter infection specific to older UK residents must be identified.