. 2019 Dec 24;9:19749. doi: 10.1038/s41598-019-55929-7

Table 2.

Characteristics and sexual practices associated with bacterial diversity of the vaginal microbiota.

Characteristic	n women reporting exposure, n intervals exposure reported (N = 100 women and 360 intervals)	Coeff. (95% CI)^a	P value^a	Adj Coeff. (95% CI)^b	P value^b
Self-reported past history of BV
No	78, 285	ref		ref
Yes	22, 75	0.28 (0.06,0.50)	0.013	0.26 (0.04,0.48)	0.018
*Longitudinal characteristics*^c
Any smoking^d,e
No	47, 210	ref		ref
Yes	53, 149	0.18 (0.01,0.35)	0.036	0.08 (−0.09,0.25)	0.352
Any douching^e
No	94, 351	ref
Yes	6, 8	0.50 (−0.02,1.01)	0.060
Number of SP
0	3, 38^f	ref
1	67, 270^f	0.03 (−0.23,0.29)	0.818
≥2	30, 52^f	0.36 (0.04,0.67)	0.027
Frequency of sex
Once/month or less	14, 118^g	ref		ref
Several times/month	36, 131^g	0.17 (−0.02,0.36)	0.074	0.13 (−0.05,0.31)	0.164
Several times/week	50, 111^g	0.30 (0.10,0.51)	0.003	0.20 (0.00,0.41)	0.049
Sex with NP^h
No	60, 300	ref		ref
Yes	40, 60	0.49 (0.30,0.68)	<0.001	0.41 (0.21,0.60)	<0.001
*Sexual practices with an FSP*ⁱ
Any receptive oral vaginal sex
No^j,e	15, 138	ref		ref
Yes	85, 221	0.24 (0.08,0.40)	0.003	0.15 (−0.02,0.31)	0.076
Any digital anal sex
No^j,e	77, 317	ref
Yes	23, 41	0.32 (0.08,0.57)	0.010
Sharing of sex toys^e
No toys/washed/condoms used^j	58, 272	ref
Unwashed	42, 87	0.18 (−0.00,0.37)	0.051
*Sexual practices with an MSP*^k
Any vaginal sex
No^{l, e}	86, 321	ref
Yes	14, 38	0.07 (−0.21,0.35)	0.637
*Self-reported symptoms*
Self-reported abnormal vaginal discharge and/or odour
No	72, 317	ref
Yes	28, 43	0.32 (0.09,0.55)	0.007

Abbreviations: BV, bacterial vaginosis; SP, sexual partner (may refer to FSP or MSP); NP, new partner (may refer to FSP or MSP); FSP, female sexual partner; MSP, male sexual partner

Missing data for variables included in this analysis occurred in <0.5% of intervals.

^aUnivariate GEE linear regression, where participant ID is panel variable. The regression coefficient represents the mean difference of Shannon diversity between the reference and comparison group/s for each characteristic/practice investigated.

^bMultivariable GEE linear regression, where participant ID is panel variable.

^cLongitudinal characteristics were measured as any exposure over the prior study interval (~90 days). No significant associations were identified between Shannon diversity and either hormonal contraceptive use or recent menses.

^dThere was no dose-response relationship between smoking and Shannon diversity.

^eMissing data from a maximum of two intervals for some variables.

^fFor women reporting different numbers of sexual partners in two or more intervals, the category representing the highest number of sexual partners has been used to calculate n women reporting exposure.

^gFor women reporting different frequencies of sexual activity in two or more intervals, the most frequent category has been used to calculate n women reporting exposure.

^hSex with a new partner with who first sexual contact was within 90 days.

ⁱNo significant associations were identified between Shannon diversity and the following sexual practices with an FSP: Any digital vaginal sex from an FSP, any receptive oral anal sex from an FSP and current FSP with BV symptoms. These practices have been left out to simplify the table.

^jOr did not have a FSP.

^kNo significant associations were identified between Shannon diversity and the following sexual practices with an MSP: Condoms used for vaginal sex, vaginal sex after anal sex, any receptive oral vaginal sex from an MSP, any digital vaginal sex from an MSP, any anal sex from an MSP. These practices have been left out to simplify the table.

^lOr did not have a MSP.