Factors associated with gonorrhoea in men aware of being positive for HIV infection: case-control study

BMJ 1998; 317 doi: (Published 17 October 1998) Cite this as: BMJ 1998;317:1052
  1. S T Sadiq (tsadi{at}, specialist registrar in genitourinary and HIV medicinea,
  2. A J Copas, research fellow in statisticsb,
  3. A M Johnson, professor of epidemiologyb
  1. Mortimer Market Centre, Camden and Islington Community NHS Trust, London WC1E 6AU
  2. Department of Sexually Transmitted Diseases, Division of Pathology and Infectious Diseases, University College London Medical School, London WC1E 6AU
  1. Correspondence to: Dr Sadiq
  • Accepted 24 February 1998

Many of those attending genitourinary medicine clinics who are aware of being positive for HIV infection are diagnosed as having new episodes of sexually transmitted diseases,1 and evidence suggests that unsafe sexual behaviour has increased among homosexual men since the early 1990s.2 Despite this, public health initiatives to reduce HIV transmission among those infected with HIV have been lacking. Assuming that gonorrhoea is a marker for risk of transmission of HIV to others,3 we conducted a case-control study among men positive for HIV attending a genitourinary clinic for regular follow up, comparing those who acquired gonorrhoea with those who did not.

Subjects, methods, and results

All cases of urethral or rectal gonorrhoea occurring among men known to be HIV positive who attended a large London HIV clinic between April 1992 and March 1996 were identified from clinic and laboratory records. Two controls infected with HIV who did not have gonorrhoea but attended the routine clinic within one day of their matched case were selected randomly from computer generated lists. For every case and matched control a period of risk began at the first attendance at the clinic or at April 1992 (whichever was later)and ended on the date of diagnosis of gonorrhoea in the case. Data were collected from case notes. Statistical analysis took account of the matching and the differing periods of risk. Odds ratios were calculated using the conditional logistic regression function of STATA.

Univariate analysis of factors associated with acquisition of gonorrhoea after adjustment for periods of risk

View this table:

All 74 cases and 145 of the 148 controls were homosexual men. The mean (SD) age of cases at date of diagnosis of gonorrhoea was 30.9 (5.5) years and of controls 37.6 (7.8). Thirty five cases had urethral gonorrhoea alone, 35 had rectal gonorrhoea alone, and 4 had both. Only age at diagnosis of HIV infection (odds ratio 0.825 (95% confidence interval 0.744 to 0.914) for every extra year) and log CD4 count (8.93 (2.91 to 27.24) for every increase by 1) were significantly associated with the acquisition of gonorrhoea in a forward stepwise model (table). These results may be partly explained by differential rates of attendance among patients and by patients attending other centres for genitourinary medicine when concerned about their sexual health.


The association between acquisition of gonorrhoea and younger age at diagnosis of HIV infection may imply a generational phenomenon or an association with age at recruitment to the study. The association with higher CD4 counts and asymptomatic disease is likely to reflect wellbeing. In our study patients infected with HIV were 10 times more likely than patients with AIDS to contract gonorrhoea. Antiretroviral treatment often results in a general improvement in wellbeing, and our findings suggest that its widespread use could lead to an increased prevalence of unsafe sex among patients infected with HIV. However, we could not test this hypothesis since such treatment became widespread after the period studied.

Patients without symptoms and those receiving current antiretroviral treatments have lower concentrations of HIV in semen,4 a probable major determinant of HIV infectivity. If such patients experience urethritis, particularly gonorrhoea,5this may offset any reduction in semen HIV concentration.

Our patients seemed not to benefit from a single session with a health adviser immediately after they had been told that they were HIV positive, though clearly a benefit may exist (table). However, bias may arise if the offer or acceptance of counselling is associated with risk behaviour. Unfortunately, documentation in the case notes was not adequate to identify such a bias. This study shows that young men and those without symptoms may constitute a key group for research into and implementation of effective interventions among HIV positive men.


We thank Mauri-Moreno, computer coordinator at the Mortimer Market Centre, for identifying cases and controls from computer databases, and Lai Cheng, senior receptionist at the centre, for collecting and collating patient notes. We also thank Patricia Maguire for verifying the data collected and the entry of data for analysis.

Contributors: STS designed the study, discussed core ideas, designed study protocols, collected all data, participated in data analysis, and drafted and revised the paper for submission. AMJ had the idea for the study, discussed core ideas, advised on the design of the study, contributed to writing the paper, and edited the paper. AJS advised on the design of the study, was responsible for data analysis, and contributed to writing the paper. Patrick French was involved in initiating the study and advised on its design. JudithStephenson advised on the study design. STS and AJS are the guarantors for the study.


  • Funding None.

  • Conflict of interest None.


View Abstract