Skip directly to site content Skip directly to page options Skip directly to A-Z link Skip directly to A-Z link Skip directly to A-Z link
Volume 26, Number 10—October 2020

Association between Shiga Toxin–Producing Escherichia coli O157:H7 stx Gene Subtype and Disease Severity, England, 2009–2019

Article Metrics
citations of this article
EID Journal Metrics on Scopus
Lisa ByrneComments to Author , Natalie Adams, and Claire JenkinsComments to Author 
Author affiliations: Public Health England, London, UK

Cite This Article


Signs and symptoms of Shiga toxin–producing Escherichia coli (STEC) serogroup O157:H7 infection range from mild gastrointestinal to bloody diarrhea and hemolytic uremic syndrome (HUS). We assessed the association between Shiga toxin gene (stx) subtype and disease severity for »3,000 patients with STEC O157:H7 in England during 2009–2019. Odds of bloody diarrhea, HUS, or both, were significantly higher for patients infected with STEC O157:H7 possessing stx2a only or stx2a combined with other stx subtypes. Odds of severe signs/symptoms were significantly higher for isolates encoding stx2a only and belonging to sublineage Ic and lineage I/II than for those encoding stx2a only and belonging to sublineage IIb, indicating that stx2a is not the only driver causing HUS. Strains of STEC O157:H7 that had stx1a were also significantly more associated with severe disease than strains with stx2c only. This finding confounds public health risk assessment algorithms based on detection of stx2 as a predictor of severe disease.

In England, infection with Shiga toxin–producing Escherichia coli (STEC) serogroup O157:H7 is relatively rare; »650 cases are reported each year (1). However, STEC O157:H7 is a pathogen of public health concern because of its potential to cause severe disease. In England, almost two thirds of case-patients reportedly experience bloody diarrhea and 5%–14% of infections progress to the severe condition of hemolytic uremic syndrome (HUS) (26).

Predictors of whether HUS will develop after STEC infection include pathogen and host factors. Most at risk for development of HUS after STEC infection are children; HUS is the leading cause of renal failure in children in developed countries, including the United Kingdom and the United States (7). Some studies have demonstrated that female sex is also associated with HUS (2,810).

The STEC pathotype is defined by the presence of the genes encoding Shiga toxin (Stx) type 1, type 2, or both, which are located on bacteriophage incorporated into the bacterial genome (11). Stx1 and Stx2 can be further divided into subtypes Stx1a-1d and Stx2a-2g. Previous studies have demonstrated an association between Stx subtype and disease severity; strains producing Stx2, particularly the Stx2a subtype, are more associated with severe disease and HUS (1216). These findings have led to the development and implementation of differential case management and public health management of cases based on Stx profile–derived STEC pathotypes in England and elsewhere (1719).

The STEC O157:H7 population has previously been delineated into 3 main lineages (I, I/II, and II) (20) and 7 sublineages (Ia, Ib, Ic, IIa, IIb, IIc, and I/II). When STEC O157:H7 emerged in England in the 1980s, the dominant lineage was I/II. Phylogenetic analyses in which hierarchical single-linkage clustering performed on pairwise single-nucleotide polymorphism (SNP) difference between strains was used revealed that almost all isolates belonging to lineage I/II fell within a 250 single-linkage SNP cluster, or clade. During the 1990s, sublineage I/II was replaced by a 250 single-linkage SNP clade within sublineage Ic (20,21). More recently, a decline in sublineage Ic and a concurrent increase in sublineage IIb have been observed (14,22). The emergence of each clade appears to coincide with the acquisition of phage encoding the stx2a gene, which, if causing more severe disease, increases the likelihood that those cases will be detected (20).

The evidence base for the differential public health management of STEC cases based on pathotype has been assimilated from relatively small studies, which prompted a review of the data in England. We therefore explored the association between Stx subtype, particularly the role of Stx2a, and disease severity in England for »3,000 cases of STEC O157:H7 reported in the 11-year period 2009–2019.

Materials and Methods

Data, Setting, and Source

For this study, we used an observational study design. In January 2009, Public Health England (PHE) implemented the National Enhanced Surveillance System for STEC (NESSS) in England. In brief, it captures standardized epidemiologic and microbiological data for all cases of STEC reported in England through an Enhanced Surveillance Questionnaire (ESQ). For each case, these data are reconciled with microbiological data in NESSS (3).

We included data on all STEC O157:H7 cases in England reported from January 1, 2009, through December 31, 2019, for which the patient submitted an ESQ and whose isolates had undergone whole-genome sequencing. For each case-patient, we extracted and coded as binary variables the following: clinical data on reported signs/symptoms (nonbloody diarrhea, bloody diarrhea, vomiting, nausea, abdominal pain, and fever); whether the patient was asymptomatic, hospitalized, or died; and whether HUS developed. We coded the responses as negative when clinical symptoms were blank on the ESQ and case-patients were not recorded as being asymptomatic. We also extracted age and sex of case-patients and categorized them as children (<16 years of age) or adults, according to a priori knowledge that children are most at risk for STEC infection and for disease progression to HUS. The outcome of interest was disease severity. Case-patients were coded as having severe disease if bloody diarrhea, HUS, or death were reported. Asymptomatic persons and case-patients with nonbloody diarrhea were considered to have mild disease. We linked data derived from whole-genome sequencing, including Stx subtype and lineage, to each case.

Laboratory Methods

In England, all fecal specimens from patients with hospital-acquired and community-acquired cases of gastrointestinal disease submitted to local hospital laboratories are tested for E. coli O157:H7. All isolates are submitted to the PHE Gastrointestinal Bacteria Reference Unit for confirmation. Since July 2015, all isolates have been sequenced for routine surveillance (National Center for Biotechnology Information Short Read Archive Bioproject no. PRJNA248042). Therefore, we included in this study all isolates received since July 2015 from case-patients with completed ESQs. In addition, we included isolates of STEC O157:H7 submitted to the Gastrointestinal Bacteria Reference Unit from January 2009 through June 2015 and sequenced as part of previous studies from case-patients with ESQs (20). The process for whole-genome sequencing has been described in detail (14).

Statistical Analyses

We used Stata 13.1 (StataCorp, for our analyses. We described cases with respect to clinically mild and severe disease by patient age, sex, and Stx subtype. We used logistic regression to investigate the relationship between Stx subtype and disease severity, adjusting for age and sex. For each variable, we calculated odds ratios (ORs) for case-patients reporting severe disease compared with those reporting mild disease. We chose the Stx2c subtype as the baseline for Stx subtype because this subtype is associated with less severe disease. To further explore the phylogenetic relationships within Stx2a isolates, we used logistic regression to investigate the relationship between Stx2a sublineages and disease severity, adjusting for age and sex. For each variable, we calculated ORs for case-patients reporting severe disease compared with those reporting mild disease.



NESSS clinical data were available for 3,241 STEC O157:H7 case-patients with genomic strain data in England during 2009–2019. Of those, 2,891 (89.2%) reported diarrheal symptoms, including 1,862 (57.5 %) who had experienced bloody diarrhea. HUS reportedly developed in 86 (2.6 %) case-patients. Thus, 1,889 (58.3%) case-patients in the dataset were categorized as having severe disease, although this proportion varied by Stx subtype (Table 1). Case-patients categorized as having mild disease accounted for 41.7 % of the dataset and included 110 asymptomatic persons. Over half (56.8 %) of case-patients in the dataset were female and 36.5 % were children <16 years of age. Severe disease was more frequently reported among female than male patients, although this difference was not significant (59.7 % vs. 56.4 %; p = 0.09), and among adults than among children (62.7% vs. 50.5 %; p<0.01).

Genomic typing data were available for isolates from 3,225 (99.5%) cases. Most (81.4%) isolates belonged to 5 specific clades within 5 sublineages: 1c (n = 789), IIa (n = 438), IIc (n = 932), I/II (n = 133), and IIb (n = 336). Infections with isolates in sublineage IIa were mostly attributed to a large outbreak associated with imported salad leaves in 2016 (23); the other 4 sublineages were associated with domestic acquisition of infection within the United Kingdom (Table 2) (24).

The dataset contained data for 86 case-patients with HUS, of which 32 were male and 54 were female. Most (66) HUS case-patients were children; infection progressed to HUS for 5.9% (66/1,119) of children, compared with 0.98% (20/2,039) of adults (Table 1).

Severity by Subtype and Multiplicative Nature

The strains of STEC O157:H7 in this dataset had genes encoding Stx1a, Stx2a, or Stx2c, or combinations of those 3 subtypes (Table 1). Of those strains that harbored 1 stx subtype, those that had stx1a or stx2a were significantly more associated with severity than those that had stx2c (Tables 13). Comparisons of the stx subtype profiles exhibited by STEC O157:H7 indicated that strains with >1 stx subtype gene are associated with higher odds of severe disease than those with 1 stx subtype gene (Table 3). When stx2c, for which disease severity was lowest, was coupled with stx1a, the odds of severity increased (OR 7.89, 95% CI 6.23–9.97) to that comparable to strains possessing stx2a only (OR 7.04, 95% CI 5.51–9.00). The highest odds of severe disease were among case-patients infected with strains harboring stx2a and stx1a (OR 19.45, 95% CI 9.20–41.16).

The most common stx profile in isolates from HUS case-patients was stx2a/stx2c (n = 50), followed by stx2a (n = 27) (Table 1). Only 2 HUS case-patients were infected with strains that did not have stx2a (both stx2c only). Five sublineages were represented among isolates from HUS case-patients: sublineage Ic (n = 54), sublineage IIa (n = 12), sublineage IIc (n = 6), lineage I/II (n = 12), and sublineage IIb (n = 2).

Subtype Stx2a is associated with 3 sublineages common in the United Kingdom: Ic, IIb, and I/II (Table 2). To explore the relationship between sublineage, clade, and severe disease, we conducted analysis by clade for Stx2a and Stx2a/2c. We found no significant difference in the odds of severity and clade for isolates encoding both Stx2a and Stx2c (Table 4). For isolates encoding Stx2a only, odds of severe symptoms were significantly higher for patients infected with isolates belonging to sublineage Ic and sublineage I/II than sublineage IIb (Table 4). Furthermore, isolates from only 2 HUS case-patients in the study belonged to sublineage IIb, despite the presence of Stx2a.


This large study of enhanced microbiological and epidemiologic data captured detailed clinical outcomes linked to molecular typing and phylogenetic analysis for adults and children infected with STEC O157:H7 in England. STEC O157:H7 is a rare but potentially very serious infection and particularly in children and elderly persons is likely to result in their interaction with healthcare services. Frontline laboratories have long had diagnostics in place and routinely screen all fecal specimens for STEC O157:H7. Therefore, NESSS captures data for a high proportion of STEC O157:H7 cases in England and is probably representative of STEC cases nationally.

Although our dataset is comprehensive, the potential for an inherent surveillance bias toward detecting more severe disease exists. Conversely, STEC HUS is underascertained in NESSS (PHE in-house data) because of challenges with the diagnosis of this condition. Moreover, for the most part, patient’s symptoms are self-reported; therefore, misclassification bias is possible, although because of the temporality of data collection, we consider bias to be low.

In our study, although HUS developed in more children <16 years of age, risk for severe disease seems to be lower than for those >16 years of age. It is possible that children are more likely to be taken for healthcare visits regardless of illness severity; therefore, our surveillance system is more likely to pick up milder cases of STEC infection in children than in adults.

Previous studies have documented the association between the presence of Stx2a and the development of HUS; thus, monitoring the presence and emergence of strains harboring this Stx subtype in the STEC population is needed (1217,19). Most of these studies included STEC from a wide variety of different serotypes, exhibiting a variety of Stx subtypes and relatively small datasets. In contrast, we analyzed a large dataset, focusing on a single serotype characterized by limited number of Stx subtype combinations. Doing so enabled us to make direct comparisons between specific Stx profiles without the confounding influence of the wide variety of virulence factors expressed by different STEC serotypes.

Our analysis revealed that the acquisition of stx1a by STEC O157:H7 also increases the association with severity. This association is significant in strains of STEC O157:H7 stx2c that acquire stx1a; the odds of severe disease from strains harboring stx1a/stx2c are comparable to the odds of severe disease from strains that have stx2a when compared with stx2c only. This finding supports previous findings that serogroups other than STEC O157 harboring stx1a only have been isolated from patients reporting severe and prolonged gastrointestinal symptoms (e.g., STEC O117) (25). Cases of bloody diarrhea and HUS caused by STEC stx1-only strains do occur, albeit at a lower frequency than cases caused by STEC harboring stx2 (26). However, the fact that stx1a-only isolates were not detected in our HUS cohort may support using presence of stx2a as a predictor of the highest likelihood of HUS development.

Analysis of the sublineages associated with HUS highlighted the rarity of sublineage IIb, despite increasing numbers of cases detected in the United Kingdom belonging to lineage IIb carrying stx2a (14). This finding correlates with the analysis showing that despite the presence of stx2a, isolates belonging to sublineage IIb are significantly less likely to be associated with severity than isolates belonging to sublineage 1c and I/II. These results indicate that the presence of stx2a is not the only driver behind HUS and that other factors are at play. These factors may include the stx-bacteriophage backbone, the stx-bacteriophage insertion site (24), copy number of the stx2a subtype gene, mutations in the stx2a subtype gene, or other gene mutations or deletions that may be involved in the expression of the toxin in vivo. A previous study (27) found that phylogenetic lineage seems to be predictive of HUS risk among those >10 years of age only and that lineage does not seem to explain HUS progression among children <10 years of age. They also observed that different lineages were observed at varying frequencies across age groups, suggestive of differences in exposure and acquisition of STEC.

This large study, which explored the association between STEC O157:H7 Stx subtype and disease severity in England over an 11-year period, provides further evidence that STEC O157:H7 exhibiting stx profiles that included stx2a only or in combination with other stx subtypes were more likely to be isolated from patients reporting bloody diarrhea, HUS, or both. However, we also observed that strains of STEC O157:H7 that had stx1a and stx2a only, or in combination with other stx subtypes, were significantly more associated with severe disease outcomes than those strains of STEC O157:H7 that had stx2c only. This finding confounds the clinical and public health risk assessment algorithms in many counties, including the United Kingdom, that are based on using detection of stx2 as a predictor of severe gastrointestinal disease.

Dr. Byrne is the lead epidemiologist for STEC at the National Infection Service, PHE, London, UK. Her research interests include public health surveillance of STEC and HUS and the investigation of outbreaks of foodborne gastrointestinal disease.



We thank Saira Butt, Bhavita Vishram, Nalini Purohit, and Mike Harte for their contributions to NESSS in England and all public health practitioners who report to the system. We also thank all the laboratory staff in the Gastrointestinal Bacteria Reference Unit, especially Michael Wright, Amy Gentle, Neil Perry, and Dawn Hedges.



  1. Public Health England. Shiga toxin-producing Escherichia coli (STEC) data: 2017 [cited 2019 Jan 31].
  2. Launders  N, Byrne  L, Jenkins  C, Harker  K, Charlett  A, Adak  GK. Disease severity of Shiga toxin-producing E. coli O157 and factors influencing the development of typical haemolytic uraemic syndrome: a retrospective cohort study, 2009-2012. BMJ Open. 2016;6:e009933. DOIPubMedGoogle Scholar
  3. Byrne  L, Jenkins  C, Launders  N, Elson  R, Adak  GK. The epidemiology, microbiology and clinical impact of Shiga toxin-producing Escherichia coli in England, 2009-2012. Epidemiol Infect. 2015;143:347587. DOIPubMedGoogle Scholar
  4. Dundas  S, Todd  WT, Stewart  AI, Murdoch  PS, Chaudhuri  AK, Hutchinson  SJ. The central Scotland Escherichia coli O157:H7 outbreak: risk factors for the hemolytic uremic syndrome and death among hospitalized patients. Clin Infect Dis. 2001;33:92331. DOIPubMedGoogle Scholar
  5. Gould  LH, Demma  L, Jones  TF, Hurd  S, Vugia  DJ, Smith  K, et al. Hemolytic uremic syndrome and death in persons with Escherichia coli O157:H7 infection, foodborne diseases active surveillance network sites, 2000-2006. Clin Infect Dis. 2009;49:14805. DOIPubMedGoogle Scholar
  6. Wong  CS, Mooney  JC, Brandt  JR, Staples  AO, Jelacic  S, Boster  DR, et al. Risk factors for the hemolytic uremic syndrome in children infected with Escherichia coli O157:H7: a multivariable analysis. Clin Infect Dis. 2012;55:3341. DOIPubMedGoogle Scholar
  7. Lynn  RM, O’Brien  SJ, Taylor  CM, Adak  GK, Chart  H, Cheasty  T, et al. Childhood hemolytic uremic syndrome, United Kingdom and Ireland. Emerg Infect Dis. 2005;11:5906. DOIPubMedGoogle Scholar
  8. Honda  T. Factors influencing the development of hemolytic uremic syndrome caused by enterohemorrhagic Escherichia coli infection: from a questionnaire survey to in vitro experiment. Pediatr Int. 1999;41:20912. DOIPubMedGoogle Scholar
  9. Byrne  L, Vanstone  GL, Perry  NT, Launders  N, Adak  GK, Godbole  G, et al. Epidemiology and microbiology of Shiga toxin-producing Escherichia coli other than serogroup O157 in England, 2009-2013. J Med Microbiol. 2014;63:11818. DOIPubMedGoogle Scholar
  10. Smith  KE, Wilker  PR, Reiter  PL, Hedican  EB, Bender  JB, Hedberg  CW. Antibiotic treatment of Escherichia coli O157 infection and the risk of hemolytic uremic syndrome, Minnesota. Pediatr Infect Dis J. 2012;31:3741. DOIPubMedGoogle Scholar
  11. Croxen  MA, Law  RJ, Scholz  R, Keeney  KM, Wlodarska  M, Finlay  BB. Recent advances in understanding enteric pathogenic Escherichia coli. Clin Microbiol Rev. 2013;26:82280. DOIPubMedGoogle Scholar
  12. Ethelberg  S, Olsen  KE, Scheutz  F, Jensen  C, Schiellerup  P, Enberg  J, et al. Virulence factors for hemolytic uremic syndrome, Denmark. Emerg Infect Dis. 2004;10:8427. DOIPubMedGoogle Scholar
  13. Persson  S, Olsen  KE, Ethelberg  S, Scheutz  F. Subtyping method for Escherichia coli shiga toxin (verocytotoxin) 2 variants and correlations to clinical manifestations. J Clin Microbiol. 2007;45:20204. DOIPubMedGoogle Scholar
  14. Byrne  L, Dallman  TJ, Adams  N, Mikhail  AFW, McCarthy  N, Jenkins  C. Highly pathogenic clone of Shiga toxin-producing Escherichia coli O157:H7, England and Wales. Emerg Infect Dis. 2018;24:23038. DOIPubMedGoogle Scholar
  15. Brandal  LT, Wester  AL, Lange  H, Løbersli  I, Lindstedt  BA, Vold  L, et al. Shiga toxin-producing escherichia coli infections in Norway, 1992-2012: characterization of isolates and identification of risk factors for haemolytic uremic syndrome. BMC Infect Dis. 2015;15:324. DOIPubMedGoogle Scholar
  16. Naseer  U, Løbersli  I, Hindrum  M, Bruvik  T, Brandal  LT. Virulence factors of Shiga toxin-producing Escherichia coli and the risk of developing haemolytic uraemic syndrome in Norway, 1992-2013. Eur J Clin Microbiol Infect Dis. 2017;36:161320. DOIPubMedGoogle Scholar
  17. Veneti  L, Lange  H, Brandal  L, Danis  K, Vold  L. Mapping of control measures to prevent secondary transmission of STEC infections in Europe during 2016 and revision of the national guidelines in Norway. Epidemiol Infect. 2019;147:e267. DOIPubMedGoogle Scholar
  18. Public Health England. Shiga toxin-producing Escherichia coli: public health management [cited 2019 Jan 31].
  19. Werber  D, Scheutz  F. The importance of integrating genetic strain information for managing cases of Shiga toxin-producing E. coli infection. Epidemiol Infect. 2019;147:e264. DOIPubMedGoogle Scholar
  20. Dallman  TJ, Ashton  PM, Byrne  L, Perry  NT, Petrovska  L, Ellis  R, et al. Applying phylogenomics to understand the emergence of Shiga-toxin-producing Escherichia coli O157:H7 strains causing severe human disease in the UK. Microb Genom. 2015;1:e000029. DOIPubMedGoogle Scholar
  21. Adams  NL, Byrne  L, Smith  GA, Elson  R, Harris  JP, Salmon  R, et al. Shiga toxin–producing Escherichia coli O157, England and Wales, 1983–2012. Emerg Infect Dis. 2016;22:5907. DOIPubMedGoogle Scholar
  22. Mikhail  AFW, Jenkins  C, Dallman  TJ, Inns  T, Douglas  A, Martín  AIC, et al. An outbreak of Shiga toxin-producing Escherichia coli O157:H7 associated with contaminated salad leaves: epidemiological, genomic and food trace back investigations. Epidemiol Infect. 2018;146:18796. DOIPubMedGoogle Scholar
  23. Gobin  M, Hawker  J, Cleary  P, Inns  T, Gardiner  D, Mikhail  A, et al. National outbreak of Shiga toxin-producing Escherichia coli O157:H7 linked to mixed salad leaves, United Kingdom, 2016. Euro Surveill. 2018;23:23. DOIPubMedGoogle Scholar
  24. Yara DA. Greig DR, Gally DL,  Dallman  TJ, Jenkins C. Comparison of Shiga toxin-encoding bacteriophages in highly pathogenic strains of Shiga toxin-producing Escherichia coli O157:H7 in the UK. Microb Genom. 2020;6:e000334.
  25. Dallman  T, Cross  L, Bishop  C, Perry  N, Olesen  B, Grant  KA, et al. Whole genome sequencing of an unusual serotype of Shiga toxin-producing Escherichia coli. Emerg Infect Dis. 2013;19:13024. DOIPubMedGoogle Scholar
  26. EFSA BIOHAZ Panel, Koutsoumanis  K, Allend  A, Alvarez-Ordóñez  A, Bover-Cid  S, Chemaly M, et al. Pathogenicity assessment of Shiga toxin-producing Escherichia coli (STEC) and the public health risk posed by contamination of food with STEC. EFSA Journal. 2020;18:5697.
  27. Tarr  GAM, Shringi  S, Oltean  HN, Mayer  J, Rabinowitz  P, Wakefield  J, et al. Importance of case age in the purported association between phylogenetics and hemolytic uremic syndrome in Escherichia coli O157:H7 infections. Epidemiol Infect. 2018;146:15505. DOIPubMedGoogle Scholar




Cite This Article

DOI: 10.3201/eid2610.200319

Original Publication Date: September 09, 2020

Table of Contents – Volume 26, Number 10—October 2020

EID Search Options
presentation_01 Advanced Article Search – Search articles by author and/or keyword.
presentation_01 Articles by Country Search – Search articles by the topic country.
presentation_01 Article Type Search – Search articles by article type and issue.



Please use the form below to submit correspondence to the authors or contact them at the following address:

Lisa Byrne, Public Health England, Gastrointestinal Infections,
61 Colindale Ave, London NW9 5EQ, UK; ; and Claire Jenkins, Public Health England, Gastrointestinal Bacteria Reference Unit, 61 Colindale Ave, London NW9 5EQ, UK

Send To

10000 character(s) remaining.


Page created: July 29, 2020
Page updated: September 17, 2020
Page reviewed: September 17, 2020
The conclusions, findings, and opinions expressed by authors contributing to this journal do not necessarily reflect the official position of the U.S. Department of Health and Human Services, the Public Health Service, the Centers for Disease Control and Prevention, or the authors' affiliated institutions. Use of trade names is for identification only and does not imply endorsement by any of the groups named above.