Table 4.

Mediation analyses of gay app use between sexuality traits and high-risk sexual behaviors of the cross-sectional study among men who have sex with men in China from August 1 to October 30, 2020^a.

Pathway	Total effect		Direct effect			Indirect effect
	cb (95% CI)	P value	c’c (95% CI)	P value	ad (95% CI)		P value	be (95% CI)	P value	a×bf (95% CI)	P value
SCg→gay app use→ MSP	0.250 (0.103-0.367)	<.001	0.227 (0.070-0.358)	.002	0.108 (0.031-0.184)		.02	0.217 (0.094-0.410)	.002	0.023 (0.004-0.067)	.01
SC→gay app use→UAIh	0.097 (0.027-0.163)	.02	0.082 (–0.002 to 0.167)	.06	0.108 (0.031-0.184)		.02	0.136 (0.056-0.205)	.003	0.015 (0.004-0.031)	.01
SSSi→gay app use→MSPj	0.241 (0.126-0.326)	.002	0.218 (0.098-0.305)	<.001	0.108 (0.014-0.205)		.03	0.218 (0.102-0.409)	.002	0.023 (0.003-0.068)	.03
SSS→gay app use→UAI	0.011 (–0.080 to 0.107)	.81	–0.005 (–0.101 to 0.094)	.93	0.108 (0.014-0.205)		.03	0.146 (0.057-0.233)	.002	0.016 (0.001-0.038)	.03

^aAll models adjust for age, marital status, education, employment, monthly income, sexual orientation, and living status.

^bTotal effect.

^cDirect effect (independent variables→dependent variables).

^dIndependent variables→mediator.

^eMediator→dependent variables.

^fIndirect effect.

^gSC: sexual compulsivity.

^hUAI: unprotected anal intercourse.

ⁱSSS: sexual sensation seeking.

^jMSP: multiple sexual partners.