Descriptive statistics
The summary of male and female incidence rates (per 100,000 populations) in different countries for each age group and relevant years is presented in Table A1, Appendix A. In every country, and in all age groups the incidence rates of Campylobacter were higher in males compared to females.
Meta-analyses by age group and group of calendar years
The meta-analyses for the age groups 0-1, 1-4, 5-9 and 10-14, are shown in Figures 1 to 4. For all age groups and countries, there was an excess in incidence rate in males. There were quantitative differences between the countries, illustrated by the relatively high I2 values. This appeared to be influenced largely by the results from Israel, where the excess in incidence rates for males was highest for most age groups. We have no explanation for this.
For infants (0-1), there was a 31% excess in the overall pooled incidence rate for males and the pooled IRRs varied from 1.21 in Germany to 1.52 in Israel (Figure 1) with overall I2 = 67%, indicating quantitative differences between the countries.
Insert Figure 1
Figure 1: Forest plot of the male to female campylobacteriosis IRR for infants (<1 year), for different years, in Canada, Germany, Israel, New Zealand, and Spain.
In the age group 1-4, there was a 34% excess in the overall pooled incidence rate in males, and the pooled IRRs varied between 22% in Germany and 51% in Israel (Figure 2) with overall I2 = 80.3%.
Insert Figure 2
Figure 2: Forest plot of the male to female campylobacteriosis IRR for early childhood (ages 1-4), for different years, in Canada, Germany, Israel, New Zealand, and Spain
In the age group 5-9, There was a 35% excess in the overall pooled incidence rate in males, with pooled IRRs varying between 21% in Germany and 47% in Israel (Figure 3) with overall I2 = 66%.
Insert Figure 3
Figure 3: Forest plot of the male to female campylobacteriosis IRR in late childhood (ages 5-9), for different years, in Australia, Canada, Finland, Germany, Israel, New Zealand, and Spain.
In the age group 10-14, there was a 73% excess the overall pooled incidence rate in males, and the IRRs varied between 47% in Finland to more than double in males in Israel (Figure 4) with overall I2 = 72.9%.
Insert Figure 4
Figure 4: Forest plot of the male to female campylobacteriosis IRR in puberty (ages 10-14), for different years, in in Australia, Canada, Finland, , Germany, Israel, New Zealand, and Spain
In the age group 15 to 44 (or 15-39) years, males had a 10% excess in incidence rates (Figure 5) The pooled IRRs varied from no excess in Germany to 19% in Spain.
Insert Figure 5
Figure 5: Forest plot of the male to female campylobacteriosis IRR in young adulthood, (ages 15-44 or 15-39), for different years, in Australia, Canada, Finland, Germany, Israel, New Zealand, and Spain
For middle-age adults (40-59 or 45-64 years), males had a 19% excess in incidence rates (Figure 6). The pooled IRRs varied from a 5% excess in Israel to a 51% excess in Spain.
Insert Figure 6
Figure 6: Forest plot of the male to female campylobacteriosis IRR in middle adulthood, (ages 45-64 or 40-59), for different years, in Australia, Canada, Finland, Germany, Israel, New Zealand, and Spain
For age 65+ or 60+ years, males had a 27% excess in incidence rates and ranged from 13% in Israel to 72% in Spain. (Figure 7)
Insert Figure 7
Figure 7: Forest plot of the male to female campylobacteriosis IRR in senior adulthood, (ages 65+ or 60+), for different years, in Australia, Canada, Finland, Germany, Israel, New Zealand, and Spain
Sensitivity analysis
To evaluate the effect of individual countries and years on the pooled IRR, we performed leave-one-out sensitivity analysis and recomputed the pooled IRRs. After omitting one country at a time, the pooled IRRs remained very similar (Table A2, Appendix A). Similar results were obtained after omission of several groups of years at a time (Table A3, Appendix A) Thus, no single country or particular groups of years substantially influenced the pooled IRRs. This confirms that the results of this study are stable and robust.
Meta-regression analysis
Meta-regression results revealed that the age groups (p<0.0001) contributed to almost all the source of heterogeneity, with very little contributed by countries or years. There was no significant difference in the pooled IRR between infancy to early/ late childhood, and senior adulthood (p>0.05).
Asymmetry analysis
Egger's test was used to check the existence of asymmetry in the contribution of countries and years in the analyses. The test was not significant for infants (p=0.427), in middle adulthood (p=0.234), and in senior adulthood (p=0.746). Evidence of asymmetry was observed for the early and late childhood, puberty and young adulthood with p <.0001 (Figure B1, Appendix B), suggesting some evidence of possible bias. We do not have any explanation for this, and it could be simply a chance finding.