Association between Semen Exposure and Incident Bacterial Vaginosis

Table 4

Results of case-crossover analyses of associations between selected factors and incident bacterial vaginosis among HIV-uninfected U.S. women, HERS, 1993–2000*.

	Control visits	Case visits	Bivariable model		Multivariable model^†
	No.	No.	OR	(95% CI)	aOR	(95% CI)
Time-dependent factors
Trichomoniasis at preceding visit
Yes	76	41	1.2	(0.7, 2.1)
No	446	189	Referent
Vaginal Candidal culture at preceding visit
Yes	143	70	1.1	(0.7, 1.6)
No	377	160	Referent
Human papillomavirus at preceding visit
Yes	126	55	1.0	(0.6, 1.5)
No	393	171	Referent
Spermatozoa detected on Gram stain
Yes	36	29	2.4	(1.3, 4.5)	2.1	(1.1, 4.0)
No	487	201	Referent		Referent
Sexual behavior during previous 6 months‡
Frequent coitus, inconsistent condom use	159	100	3.3	(1.7, 6.6)	3.0	(1.5, 5.9)
Frequent coitus, consistent condom use	46	29	3.1	(1.3, 7.3)	3.1	(1.3, 7.4)
Infrequent coitus, inconsistent condom use	114	36	1.7	(0.8, 3.4)	1.6	(0.8, 3.2)
Infrequent coitus, consistent condom use	58	19	1.1	(0.5, 2.4)	1.1	(0.5, 2.5)
Not sexually active	145	45	Referent		Referent
Female sex partner during previous 6 months
Yes	48	25	1.7	(0.6, 4.6)
No	472	204	Referent
Douching within 48 hours
Yes	23	10	0.8	(0.3, 2.2)
No	497	220	Referent
Current hormonal contraception use
Yes	22	8	0.4	(0.1, 1.3)
No	500	221	Referent
Current injection drug use
Yes	93	42	1.5	(0.7, 3.2)
No	429	188	Referent
Crack use during previous 6 months
Yes	93	45	1.4	(0.8, 2.5)
No	429	185	Referent
Cigarette use during previous 6 months
Yes	373	173	1.4	(0.6, 3.5)
No	149	57	Referent

*Findings from conditional logistic regression model based on 230 case visits (i.e., visits with incident bacterial vaginosis) and 523 control visits (i.e., visits without incident bacterial vaginosis) completed by 159 women with at least one case and one control visit.
^†Adjusted for all variables in column.
^‡Frequent coitus defined as ≥4 times per month and infrequent coitus defined as <4 times per month.