RR:C19 Evidence Scale rating by reviewer:
Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.
***************************************
Review: Anthony Ebert and colleagues examined the risk of inflammatory bowel disease (IBD) following hospitalization with infectious mononucleosis (IM), in a nationwide Danish cohort study between 1977 and 2021. The group utilized the Danish National Civil Register (CPR) and the Danish National Patient Register (LPR). They identified 39,684 patients with severe IM who were sex-, age-, and municipality-matched to 396,840 non-IM hospitalized controls. “Severe IM was significantly associated with the development of IBD (HR:1.35; 95% CI: 1.22-1.49) and this was seen particularly in CD (HR: 1.56; 95% CI: 1.34-1.83)”. The study is valuable in providing plausible support for EBV participating in the infectious environmental origins of IBD. However, as with all epidemiologic studies, hidden co-variants (hidden bias) are likely to be present and should be emphasized. We recommend addressing the following questions:
It is unclear how the diagnoses of EBV infection/IM were made. There are numerous infectious and autoimmune diseases that have been recognized to associate with false positive IgM based testing for EBV (such as the monospot), for example. Those people who are prone to develop IBD are more likely to generate non-specific exaggerated immune responses during other infections and illnesses (which can lead to such false negative IgM elevations during such illnesses) compared to those people who are protected against immune medicated disorders.
Other hospitalized viral infectious controls could significantly increase the value of this study since the investigators only examined Chlamydia trachomatis infection as a control (it is unclear from the manuscript whether these infectious cases were also hospitalized, or not). The fact that severe IM cases did not have increased prevalence of subsequent IBD compared those with Chlamydia trachomatis infection already raises concerns for the EBV related findings to be non-specific (it is interesting that the findings on Chlamydia trachomatis infection are not discussed in the Discussion). Comparisons with severe CMV or Influenza infections and subsequent IBD could address this concern.
The time from IM hospitalization to the first diagnosis of IBD is not detailed. The implications for immediate or near IBD diagnosis IM (i.e. “trigger”) versus several years elapsing between IM and subsequent IBD (i.e. “IBD prone” state) are much different in regards to the developmental origins of IBD (see R. Kellermayer , M. Zilbauer https://pubmed.ncbi.nlm.nih.gov/33093364/). Crohn’s disease (CD) diagnosis, for example, can be delayed by several months; therefore, it is possible that several hospitalized IM cases may have had CD already. Comparison between IBD diagnoses within 1 year from IM hospitalization versus more than 1 year is recommended.
The diagnosis of severe IM resulting in a hospital diagnosis may not truly be reflective of the severity of the course of IM (i.e. did this include emergency room [ER] diagnosis as 0 days of admission was included in the admission range, however, practically speaking less than one day of admission seems to suggest the course was not truly severe). Could this factor have influenced the results of the Koge hospital population? At least based on this sub-population, one cannot conclude that it is the severity of IM (i.e., hospitalization versus not) that determines the association between EBV infection and subsequent IBD development (since hospitalized IM cases were not more commonly associated with IBD than not hospitalized cases). It is interesting to note that the findings of the Koge hospital cohort are not discussed in the Discussion. Consequently, the conclusion “The main findings of this study indicate that hospitalization with IM is associated with a significantly increased risk of IBD development.” can be seriously questioned (see point 2 as well).
The study was performed in one of the economically most advanced countries where IBD has one of the highest incidences. In such countries, EBV infection and IM shifts to older children in whom the severity of the infection is higher making IM severity and IBD prevalence dependent variables (i.e. both associated with economic advancement/industrialization). Therefore, the findings of this study may not be applicable to economically less advanced areas of the world, which is not highlighted in the discussion. Additionally, association vs. causation between EBV/IM and IBD can be further questioned.
Minor:
While the authors emphasize the reliability of their methodology for IBD diagnosis based on the Danish registries (reference 42 and 43), they forgot to recognize that the actual diagnosis can be questioned in over 7% of the patients due to the lack of appropriate diagnostic evaluation preceding the diagnosis of IBD (see NF Rasmussen, et al. https://pubmed.ncbi.nlm.nih.gov/36028261/)
The methods for identifying individuals hospitalized (?) with Chlamydia trachomatis infection should be detailed in the methods
Creative Commons Attribution 4.0 International License (CC-BY 4.0)
