Comparison with previous studies
Several population-based studies, including a large meta-analysis, have reported an increased incidence and prevalence of EoE in Europe and North America over time (9, 10). Recently, the meta-analysis including over 288 million participants and 147668 patients with EoE from 15 countries across the five continents was published. The global pooled incidence and prevalence of EoE were 5.31 cases per 100,000 inhabitant-years (95% CI, 3.98–6.63; number of studies, 27; sample population, 42,191,506) and 40.04 cases per 100,000 inhabitant-years (95% CI, 31.10–48.98; number of studies, 20; sample population, 30,467,177), respectively. The pooled incidence of EoE was higher in high-income countries (vs low- or middle-income countries), males, and North America (vs Europe and Asia). The global prevalence of EoE followed a similar pattern. The pooled prevalence of EoE gradually increased from 1976 to 2022 (1976–2001; 8.18; 95% CI, 3.67–12.69 vs 2017–2022; 74.42; 95% CI, 39.66–109.19 cases per 100,000 inhabitant-years) (20).
One of the possible explanations is higher disease awareness. The number of publications related to EoE increased substantially from 259 publications per year in 2012 to 418 publications per year in 2021.
The previous Dutch (21) and Swedish studies (11) used esophageal biopsy reports to identify EoE patients. Nationwide Dutch study reported a remarkable increase of EoE incidence over 25 years, in where the incidence was 0.01 per 100,000 in 1995 and 3.16 per 100,000 inhabitants in 2019 (21), which is similar to our findings (4.42 per 100,000 inhabitants in 2019). The Swedish study between 2004 and 2015 reported an overall incidence of 1.22 per 100,000 person years with an increase of 33% on an annual basis which is twice as much as we observed during our study period (average 15% per year) (11). However, the authors did not disclose any information regarding where in the esophagus the biopsies were obtained, which may have resulted in an overlap of reflux patients in the study cohort.
Our study was based on ICD-10 codes from the National Patient Registry, similar to the approach applied by Delon et al in a Danish cohort in 2015 reporting 20-fold increase of the incidence rate between 1997 and 2012 (12). In our cohort, EoE was more common in men (almost a 3:1 ratio), and in the third decade of life, which also corroborates with previous findings (11, 12, 21).
Our study reports geographical variations of the incidence of prevalence of EoE within Sweden during a specific study period, demonstrating an increase of EoE in 19 out of 21 regions. In one of the largest regions in Sweden (Västra Götaland) the incidence has been approximately four per 100,000 person years during the entire study period. Geographic risk factors explaining the regional differences, have previously been reported in the United States. Several studies have utilized large nationwide pathology databases containing data for hundreds of thousands of patients undergoing endoscopy with esophageal biopsy, representing more than 14,000 cases of esophageal eosinophilia and EoE. Prevalence of EoE was noted to vary by climate zone, with cases more commonly noted in arid and cold weather climates (22, 23).
Furthermore, differences in EoE management across Europe have been reported (24), which could also explain our findings within Sweden. Regions with highest prevalence of EoE in Sweden, Västra Götaland and Stockholm- Gotland, are geographic catchment areas for tertiary referral institutions. These regions, together with Skåne, are covering more than 53% of the Swedish population (25).
It is possible that the level of interest and awareness of EoE among clinicians and pathologists in the regions with high prevalence may not reflect the disease awareness in other regions.
The awareness of EoE in Sweden has been low until recently. For example, in 2011 no record of EoE was observed. The European EoE guidelines were published in 2011 (26) and updated in 2017 (27). The Swedish national guidelines were published in 2021 (28). In addition to the late introduced guidelines, compared to our study period, the availability of endoscopic examinations differs between the regions (29). For example, 23% of the patients in Halland were examined beyond the 90-day health care guaranty, compared to 70% of the patients in Norrbotten (29).
Within European countries estimated median diagnostic delay was 5.0 (1,9 − 8,3) years from 2008 to 2011, 2,8 (0,7 − 5,2) years from 2012 to 2013 and 1,2 (0,5 − 2,9) years from 2014 to 2017 (9). Recently published studies from Denmark found that in Danish EoE patient cohort diagnostic delay was initially higher than in other European countries giving a lack of EoE awareness among physicians in Denmark as an explanation to the findings (30). Diagnostic delay in 2007 to 2017 was 5,5 (2,0;12) years and decreased to median of 4 years (1;10) from 2018 to 2021 (31).
Previous studies have suggested changes in hygiene conditions (32), diet and lifestyle factors (33), and environmental factors (34), as other possible reasons for the increased incidence. During the study period, the proportion of patients treated with PPI decreased. The introduction of new treatment alternatives and discarding the previously mandatory PPI trial might be an explanation. PPI responsive eosinophilia is not a separate diagnosis that should be excluded and patients with PPI responsiveness should be considered within the spectrum of EoE (35).