Table 3.
Multivariable Cox regression analysis of HIV incidence among men who have sex with men retained in our cohort in Shenyang (N=461).
| Characteristics of users and nonusersa | Crude analysis | Multivariable analysisb, cHRc (95% CI) | ||
| aHRd (95% CI) | P value | |||
| Age of sexual debut (years) | ||||
| ≤20 | 1.3 (0.5-3.3) | 1.1 (0.4-2.8) | .90 | |
| >20 | Reference | Reference | —e | |
| Ever had sexual intercourse with females | ||||
| Yes | 0.5 (0.2-1.3) | 0.6 (0.2-1.6) | .28 | |
| No | Reference | Reference | — | |
| Anal sex position | ||||
| Versatile | 0.4 (0.1-1.2) | 0.5 (0.1-1.6) | .24 | |
| Bottom | 0.5 (0.1-1.4) | 0.5 (0.1-1.7) | .27 | |
| Top | Reference | Reference | — | |
| Ever been tested for HIV | ||||
| Yes | 1.5 (0.5-4.1) | 1.2 (0.4-3.4) | .75 | |
| No | Reference | Reference | — | |
| Used recreational drugs in the past 3 months | ||||
| Yes | 2.5 (1.0-6.2) | 2.6 (1.0-6.9) | .048 | |
| No | Reference | Reference | — | |
| Number of male SPsf in the past 3 months | ||||
| ≥2 | N/Ag | N/A | N/A | |
| <2 | N/A | N/A | — | |
| Number of male CPsh in the past 3 months | ||||
| ≥5 | 1.7 (0.5-6.0) | 1.9 (0.6-6.7) | .30 | |
| <5 | Reference | Reference | — | |
| CAIi with male CPs in the past 3 months | ||||
| Yes | 2.9 (1.1-7.4) | 3.2 (1.2-8.4) | .02 | |
| No | Reference | Reference | — | |
| CAI with male SPs in the past 3 months | ||||
| Yes | 0.3 (0.1-1.4) | 0.3 (0.1-1.4) | .12 | |
| No | Reference | Reference | — | |
| Had group sex with males in the past 3 months | ||||
| Yes | 4.3 (1.4-13.1) | 4.8 (1.6-15.0) | .01 | |
| No | Reference | Reference | — | |
| Positive for syphilis at baseline | ||||
| Yes | 1.2 (0.4-4.1) | 1.3 (0.4- 4.8) | .67 | |
| No | Reference | Reference | — | |
| Ever used GSNj apps to seek male sexual partners | ||||
| Yes | 4.1 (1.2-14.2) | 3.7 (1.1-13.1) | .04 | |
| No | Reference | Reference | — | |
| Used GSN apps to seek male sexual partners in the past 3 months | ||||
| Yes | 2.9 (1.0-7.9) | 2.6 (0.9-7.9) | .08 | |
| No | Reference | Reference | — | |
aThe number of GSN apps’ users and nonusers retained to at least one 3-month follow-up visit.
bThe multivariable Cox regression model adjusted for age, residence status, ethnicity, education, income, and marital status.
ccHR: crude hazard ratio.
daHR: adjusted hazard ratio.
eNot applicable.
fSPs: steady partners.
gN/A: not available.
hCPs: casual partners.
iCAI: condomless anal intercourse.
jGSN: geosocial networking.