Umeå universitets logga

Background: Waterborne infection with the parasite Cryptosporidium is a common cause of diarrhoea, particularly inlow-income countries, and contributes significantly to diarrhoea-related morbidity and mortality. Young children andimmunocompromised individuals are particularly vulnerable to such infections. In 2010, an outbreak of Cryptosporidiumhominis in Östersund, Sweden, resulted in ~27,000 cases. Long-term symptoms, including abdominal and joint symptoms,headache, and fatigue, can persist for up to 5 years, but whether they extend beyond that time frame is unclear. Data on thelong-term health-economic impact of Cryptosporidium outbreaks are lacking. In addition, physicians appreciate that theincidence of inflammatory bowel disease (IBD) and microscopic colitis (MC) has increased in Region Jämtland Härjedalen(RJH) since 2010, but a link between cryptosporidiosis and IBD or MC has not been established.

Aim: We aimed to understand the long-term consequences of the Crypto-sporidium outbreak in Östersund by assessingsymptoms after 10 years, as well as the health-economic impacts and incidences of IBD and MC.

Methods: In 2011, a randomly selected cohort of Östersund residents (n = 1,524) and 500 additional children aged 0–5years were surveyed via a postal questionnaire. Respondents reporting new episodes of diarrhoea during the outbreak weredefined as cases; those who did not report diarrhoea were defined as non-cases. In a prospective cohort study conducted in2021, all respondents (n = 1,404, 69.4%) were surveyed about possible long-term symptoms. We used logistic regressionanalysis to examine a possible association between reported symptoms and case status. The results were publishedseparately for adults (Study I) and children aged 0-5 years (Study II). For adults, we evaluated the consistency of reportedsymptoms, including the results of a 5-year follow-up of the same cohort, and its possible association with disease durationduring the outbreak. For children, we investigated healthcare utilisation and diagnoses during the follow-up period.All questionnaire respondents in 2011 were included in a retrospective, incremental cost analysis focusing on productionloss and healthcare utilisation during 2011-2019 (Study III). Data were collected on net days of sickness and childcarebenefits, outpatient visits, and inpatient days. Possible differences between cases and non-cases were assessed by logisticand negative binomial regression analyses. Costs were calculated using population data and mean wages.In an observational study (Study IV), we compared pre-outbreak (2006-2010) and post-outbreak (2011-2019) incidences ofIBD and MC in RJH using negative binomial regression analysis.

Results: In 2021, 538 adults (74%) responded to the questionnaire. Cases were more likely to report abdominal (aOR ~3)and joint symptoms (aOR ~2). They were also more likely to report symptoms at both the 5- and 10-year follow-up. Caseswith consistent abdominal symptoms had more symptomatic days during the outbreak than the cases reporting no orvarying symptoms (mean 9.2 days [median 6.5, range 1.0-55.0] vs 6.6 days [median 5.0, range 0.0-33.0]; p = 0.003).Among children aged 0-5 years, 295 (69.7%) responded. Cases were more likely to report joint symptoms (aOR 4.0, 95% CI1.3–12.0) and fatigue (aOR 1.9, 95% CI 1.1–3.4), although overall numbers were low. No between-group differences wereobserved in healthcare utilisation or disease diagnoses.The cost analysis revealed that more cases than non-cases received sickness benefits in the years following the outbreak(142/346 [41.0%] vs 125/399 [31.3%], p = 0.006). Cases also claimed more benefit days (8,946 per 100 cases vs 4,538 per100 non-cases; p = 0.003). This corresponds with an incremental cost of ~ 33.3 million SEK (€2.9 million) for our studypopulation, or ~ 1.7 billion SEK (€148.5 million) for the entire population of Östersund. No differences were observed inchildcare benefits or outpatient visits. Hospital admission rates were comparable between groups, though cases generallyhad shorter hospital stays, likely reflecting factors unrelated to cryptosporidiosis.In RJH, 410 patients were diagnosed with IBD and 155 with MC between 2006 and 2019. For the population >40 years old,the incidence of IBD (IRR 1.69, 95% CI 1.13–2.51) and the incidence of Crohn’s disease in particular (IRR 2.23, 95% CI1.08–4.62) increased post-outbreak. The incidence of MC increased sixfold (IRR 6.43, 95% CI 2.78–14.87).

Conclusion: Long-term symptoms commonly occur following a C. hominis infection, particularly in adults. The 2010outbreak generated an estimated 1.7 billion SEK increase in sickness benefits. Furthermore, post-outbreak, the incidence oflate-onset IBD and MC increased. These findings underscore the importance of prevention and may guide future public health interventions and water management policies.