Editor,—We are puzzled by the surprisingly little, if any, serious work done to explain the epidemiological enigma of high prevalence of bacterial vaginosis (BV) in lesbians,¹ and the oft observed, but as yet unconfirmed association between BV and receptive cunnilingus in women in general.

In a detailed study of 17 consecutive lesbians attending the department of genitourinary medicine at the Royal Sussex County Hospital in Brighton, bacterial vaginosis was found in six women (35%). Of nine lesbians who practised receptive cunnilingus in the previous 4 weeks, six (67%) had BV. By contrast, no BV was present in all eight women who did not practise oral sex (table 1).

In a parallel prospective study of 256 consecutive heterosexual female patients attending the same department, 55 (21%) were diagnosed as having BV. Of 111 women who practised receptive cunnilingus in the previous 4 weeks, 41 (37%) had BV. Of 145 women who did not have oral sex, only 14 (10%) had BV (table 1). In both groups there was strong association between BV and receptive cunnilingus (p<0.001).

The evidence associating bacterial vaginosis with oral sex is too strong to be ignored and repeatedly dismissed. The mouth is full of Gram positive and Gram negative organisms including Bacteroides oralis and, albeit in much smaller quantities, lactobacilli. These organisms are part of normal flora in the mouth, but are they normal to the vagina? Might the tiny amount of lactobacilli be enough to act as a phage which destroys the endogenous healthy vaginal lactobacillus? In an interesting hypothesis, Blackwell described the possible effect of biochemical and microbial abnormalities in the vagina on BV recurrence.² She also quoted Berger's description of concordant vaginal floras in lesbian couples, suggestive of a mechanical transfer of an infectious agent.³ Is it not possible for mouth organisms or hostile salivary enzymes to induce biological and microbial abnormalities in the vagina?

Furthermore, mechanical transfer of infectious agents in lesbian couples is most likely to occur via cunnilingus, a not uncommon practice among lesbians.

Cunnilingus is a common fact of sexual life. The dynamics of this practice vary considerably. If association between BV and oral sex is ever confirmed, would the degree of tongue penetration be a factor and should it be incorporated in the aetiology equation? Further and more extensive studies are certainly indicated.