Article Text

Download PDFPDF

Short report
Shigellosis outbreak among MSM living with HIV: a case–control study in Taiwan, 2015–2016
  1. Hao-Hsin Wu1,
  2. Yi-Ting Shen2,
  3. Chien-Shun Chiou3,
  4. Chi-Tai Fang4,5,
  5. Yi-Chun Lo6
  1. 1 Division of Infection Control and Biosafety, Centers for Disease Control, Taipei, Taiwan
  2. 2 Office of Preventive Medicine, Centers for Disease Control, Taipei, Taiwan
  3. 3 Center for Diagnostics and Vaccine Development, Centers for Disease Control, Taipei, Taiwan
  4. 4 Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, Taipei, Taiwan
  5. 5 Division of Infectious Diseases, Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan
  6. 6 Office of Deputy Director, Centers for Disease Control, Taipei, Taiwan
  1. Correspondence to Professor Chi-Tai Fang, Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, 17 Xuzhou Road, Taipei 100, Taiwan; fangct{at} and Dr Yi-Chun Lo, Centers for Disease Control, Taipei 100, Taiwan; loyichun{at}


Objectives Outbreaks of shigellosis among men who have sex with men (MSM) have been reported since the late 1990s. HIV infection is an important risk factor. Since 2014, the global shigellosis epidemic has intensified. Whether chemsex (the use of crystal methamphetamine, γ-hydroxybutyrate or mephedrone to enhance sex) is a new risk factor has not been previously examined.

Methods We conducted a population-based, case–control study in Taiwan. Acute shigellosis cases diagnosed during the 2015 outbreak among MSM living with HIV were compared with those without shigellosis. CD4+ counts, plasma viral load (pVL), gonorrhoea, syphilis and amoebiasis records were obtained from the Notifiable Disease Surveillance System database. We invited cases/controls to provide information on illicit drug use and sexual behaviours, using a structured questionnaire.

Results Seventy-five shigellosis cases were compared with 225 controls. High pVL (>100 000 copies/mL; adjusted OR (aOR): 4.9, 95% CI 1.4 to 16.9), gonorrhoea (aOR: 29.4, 95% CI 2.3 to 340.2) and syphilis (aOR: 4.3, 95% CI 1.6 to 11.6) were independent risk factors of shigellosis. Twenty shigellosis cases and 59 controls completed the questionnaire. Oral-to-anal sex (aOR: 15.5, 95% CI 3.6 to 66.7), chemsex (aOR: 5.6, 95% CI 1.4 to 22.7) and poppers use (aOR: 10.9, 95% CI 1.9 to 64.2) within 12 months were independent behavioural risk factors of shigellosis.

Conclusions Chemsex is a new risk factor for shigellosis among MSM living with HIV, as identified in the 2015–2016 outbreak. Additional risk factors include poppers use, sexual risk behaviours and high pVL. Further studies on chemsex among MSM, which is a rising public health concern, are urgently required.

  • HIV
  • bacillary dysentery
  • methamphetamine
  • sodium oxybate
  • amyl nitrite
  • sexual behaviour

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Shigella spp., which cause bacillary dysentery, can be transmitted through oral-to-anal sex between men.1 Outbreaks of shigellosis among men who have sex with men (MSM) have been noted since the late 1990s,2 with HIV infection reported to be an important risk factor.2 3 Probable mechanisms include both behavioural cofactors and biological susceptibility.4

The sexually transmitted shigellosis epidemic among MSM has intensified since 2014,5 6 with more than half of new cases occurring in people using drugs, typically crystal methamphetamine, γ-hydroxybutyrate (GHB) or mephedrone, to enhance sex. This new trend was first reported in 2013 and is now termed ‘chemsex’.7 8 However, chemsex as a new risk factor for shigellosis has not been investigated.

Our previous molecular epidemiological analyses of Shigella isolates from the 2015 outbreak among MSM in Taiwan showed that the Taiwanese outbreak was part of a global shigellosis epidemic.6 9 In the present case–control study, we aimed to examine the role of chemsex in the 2015–2016 shigellosis outbreaks among MSM.


This was a nationwide, population-based, case–control study.

Shigellosis has been a notifiable disease in Taiwan since 1999. Healthcare providers need to report all culture-positive shigellosis cases through the Notifiable Disease Surveillance System (NDSS) within 24 hours after diagnosis, and submit Shigella isolates to the Taiwan Centers for Disease Control for laboratory confirmation.

We identified all domestic shigellosis cases that occurred during 2015–2016 among people living with HIV from the NDSS, HIV infection also being a notifiable condition. Each HIV-positive shigellosis case was individually matched to three HIV-positive controls without shigellosis by age (±5 years), sex, HIV diagnosis date (±30 days), county/city of residence and survival status.

Shigella species identification and antimicrobial susceptibility testing used in this study were based on the standard method.6 We obtained data from the NDSS on demographics, HIV risk factors (MSM vs other risk behaviours), most recent CD4+ count and plasma viral load (pVL) before the onset of shigellosis, antiretroviral therapy (ART), and outpatient visits for HIV care, as well as history of syphilis, gonorrhoea and Entamoeba histolytica infection (EHI).

To obtain behavioural information, a trained public health nurse contacted cases diagnosed with shigellosis in 2015 and their controls between April and September 2016. At least three attempts were made to contact each case/control. The nurse interviewed those who agreed to participate by telephone, using a structured questionnaire.

Chemsex was defined as intentional sex under the influence of crystal methamphetamine, GHB, γ-butyrolactone (GBL) or mephedrone immediately proceeding or during the sexual session.8

Statistical analyses were performed using SAS V.9.4. All variables with a P value <0.10 on univariate analysis were included in the maximum model and underwent stepwise selection during multivariate analyses. A P value <0.05 was considered statistically significant.


From January 2015 to July 2016, 79 of the 147 shigellosis cases occurred in people living with HIV. The Shigella isolates were S. sonnei in 54 cases and S. flexneri in 25 cases. Of the 71 isolates for which antimicrobial susceptibility was available, S. sonnei were more likely to be resistant to ciprofloxacin (44/47 (94%) vs 2/24 (8%); P<0.001) and trimethoprim/sulfam ethoxazole (37/47 (79%) vs 2/24 (8%); P<0.001), while S. flexneri were more likely to be resistant to ampicillin (22/24 (92%) vs 5/47 (11%); P<0.001).

In contrast, there were no differences in epidemiological characteristics between S. sonnei and S. flexneri among the 75 shigellosis cases in MSM living with HIV (age <40: 22/52 vs 6/23; CD4+ count <0.35×10^9/L: 20/52 vs 7/23; pVL >100 000 copies/mL: 12/52 vs 3/23; on/ever used ART: 21/52 vs 15/23; gonorrhoea: 10/52 vs 9/23; syphilis within 6 months: 13/52 vs 4/23; syphilis >6 months: 31/52 vs 16/23; and EHI: 3/52 vs 2/23, respectively; all P>0.05).

Seventy-five MSM shigellosis cases with available CD4+ count and pVL data were successfully matched to 225 MSM controls living with HIV. Compared with controls, shigellosis cases had a significantly lower CD4+ count, a higher pVL, were less likely to start ART or attend HIV outpatient clinics within 12 months, and were more likely to have past syphilis or gonorrhoea infection. Multivariate conditional logistic regression analyses showed that a pVL ≥100 000 copies/mL (adjusted OR (aOR): 4.9, 95% CI 1.4 to 16.9), past gonorrhoea infection (aOR: 29.4, 95% CI 2.3 to 340.2), syphilis infection within 6 months (aOR: 4.3, 95% CI 1.6 to 11.6) and past syphilis infection from >6 months (aOR: 3.3, 95% CI 1.5 to 7.0) were independent risk factors for shigellosis.

We successfully contacted 22 of the 39 shigellosis cases diagnosed in 2015, and 65 of their 195 matched controls, with 20 cases and 59 controls subjects agreeing to complete the questionnaire. Table 1 shows the demographic and sexual behavioural data. More than half of the cases (13/20, 65%) had engaged in chemsex within 12 months before the diagnosis of shigellosis.

Table 1

Univariate and multivariate analyses for behavioural risk factors of shigellosis among 79 MSM living with HIV in Taiwan, 2015–2016

Compared with controls, shigellosis cases had more sexual partners, were more likely to seek casual sex partners online, were more likely to use crystal methamphetamine, popper, sildenafil/tadalafil or marijuana, and were more likely to engage in oral-to-anal sex or chemsex within 12 months (table 1).

Multivariate logistic regression analyses showed that oral-to-anal sex (aOR: 15.5; 95% CI 3.6 to 66.7), chemsex (aOR: 5.6; 95% CI 1.4 to 22.7) and poppers use (aOR: 10.9; 95% CI 1.9 to 64.2) within 12 months were independent risk factors for shigellosis (table 1).


Our results show that chemsex is a new risk factor for sexually transmitted shigellosis among MSM, independent from HIV infection. We also identify poppers use, past STIs (gonorrhoea, syphilis) and a high plasma HIV viral load (>100 000 copies/mL) as additional risk factors.

The present investigation is the first nationwide, population-based, case–control study on the 2015 shigellosis outbreak among MSM in Taiwan. Since the majority of shigellosis cases in this outbreak occurred in people living with HIV, we focused on shigellosis cases among MSM living with HIV, and enrolled MSM living with HIV without shigellosis as controls. The strength of this approach is the enhanced statistical efficiency in identifying new risk factors behind this outbreak, by matching cases and controls based on HIV infection, which is a well-established risk factor for shigellosis among MSM.3

Chemsex, which was first described in England,7 is a relatively new public health concern worldwide.7 Unlike older illicit drugs, crystal amphetamine, GHB/GBL and mephedrone appear to make people lose their inhibition and simultaneously increase their sexual durability. Although oral-to-anal sex, or ‘rimming’, is not a new practice, a chemsex-associated increase in numbers and duration of sexual contacts could facilitate an outbreak of shigellosis, or worsen an existing epidemic.

Poppers, an anal sphincter relaxant, also promote sexual desire, reduce sexual inhibition and prolong the duration of sexual activities. Previous studies have shown that, similar to chemsex, poppers use is associated with an increase in sexual risk behaviours among MSM.10 Little is known regarding the social context behind the now-increasing use of chemsex and poppers in Taiwan. A well-conducted qualitative study will help to obtain important information and provide indepth insight into this complex issue.

In addition to the behavioural risk factors noted above, biological susceptibility from HIV-associated immunosuppression is a proposed mechanism explaining the link between HIV infection and shigellosis in MSM.3 4 In the present study, high pVL was found to be an independent risk factor for shigellosis after adjusting for behavioural risk, as measured by STIs. Our findings provide new evidence to support the proposed role of biological susceptibility in shigellosis outbreaks among MSM living with HIV. This also suggests that commencing ART early may reduce the risk of shigellosis within this population, although further studies are required to confirm this hypothesis.

Our study has several important limitations. First, information on sexual risk behaviours and illicit drug use could be under-reported. Second, only half (20/39) of candidate cases and one-third (59/179) of candidate controls were willing to provide behavioural information. The limited sample size did not allow for adequate cross-analyses of behavioural and clinical data. Finally, our study focused on MSM living with HIV, and the results might not be applicable to MSM not infected with HIV.

The difference in antibiotic resistance profiles between S. flexneri and S. sonnei isolates poses a challenge when clinicians choose an antibiotic for an HIV-positive patient with suspected shigellosis. Healthcare providers are now advised to order a stool culture to obtain isolates for susceptibility testing.

In conclusion, the present study shows that chemsex was a new risk factor for shigellosis in the 2015–2016 outbreak. Additional risk factors include poppers, sexual risk behaviours and high plasma HIV viral load. Further studies on the emerging practice of chemsex among MSM, which remains undersurveyed in Asia, are urgently required.


We would like to thank Uni-edit ( for copyediting this manuscript.



  • H-HW and Y-TS contributed equally.

  • C-TF and Y-CL contributed equally.

  • Handling editor Joseph D Tucker

  • Contributors C-TF and Y-CL conceived and designed the study. H-HW conducted the matched case–control analyses on the roles of CD4, pVL, past gonorrhoea, syphilis and amoebic infections. Y-TS enrolled the cases and controls through telephone interview, obtained information on illicit drug use and sexual behaviours, and conducted the unmatched case–control analyses on the roles of chemsex, poppers use and oral-to-anal sex. C-SC conducted laboratory confirmation of Shigella isolates, species identification and performed antimicrobial susceptibility testing. H-HW wrote the first manuscript. C-TF wrote the revised manuscript. H-HW, Y-CL, C-SC and Y-TS read and approved the final submitted version.

  • Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Ethics approval The study procedures were approved by the Institutional Review Board of Taiwan Centers For Disease Control (TCDC: registration numbers 1-04 301 and 105 303).

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Correction notice This paper has been amended since it was published Online First. There are now two corresponding authors and we have updated the paper accordingly.