• Abnormal vaginal bleeding;
  • cancer;
  • epidemiology;
  • intermenstrual bleeding;
  • postcoital bleeding


To obtain estimates of the rates of occurrence and spontaneous resolution of intermenstrual and postcoital bleeding, and investigate any association with underlying malignancy.


Two-year prospective cohort study with medical record review during the survey period, and for the subsequent 2 years.


Seven general practices with 67 100 registered patients.


All women aged 40–54 years on the practices age–sex registers.


Baseline postal questionnaire, with follow-up questionnaires sent to naturally menstruating respondents at 6, 12, 18 and 24 months. Medical record review using computerised searches from baseline to 48 months.

Main outcome measures

Prevalence and incidence of intermenstrual and postcoital bleeding, and rate of spontaneous resolution.


A total of 7121 baseline questionnaires were sent out, with an initial response rate of 66%. A total of 2104 naturally menstruating women were recruited for the prospective cohort study. The 2–year cumulative incidence of intermenstrual bleeding was 24% (95% CI 21–27%), and that of postcoital bleeding was 7.7% (95% CI 6.2–9.5%). The rates of spontaneous resolution without recurrence for 2 years were 37% (95% CI 29–45) and 51% (95% CI 40–62), respectively. Of the 785 women identified with intermenstrual and/or postcoital bleeding, only one developed uterine cancer.


There is a high prevalence, incidence, and spontaneous rate of resolution of intermenstrual and postcoital bleeding in naturally menstruating women during the perimenopausal years. The association of these symptoms with malignancy is weak. This is of importance to women in deciding when to consult and to those devising public health campaigns concerning symptoms of possible oncological significance.