Gender disparities in the mediating role of symptom knowledge level in reducing acute coronary syndrome (ACS) decision delay: Findings from a community-based study in China

Table 3 Mediation effect on decision delay stratified by gender

Exposure	Gender	Model 1^a	Model 2^b	Model 3^c
Exposure	Gender	Total effect (OR,95%CI, P)	Assoc. Mediator (β,95%CI, P)	Mediation effect (%, P)	Direct effect (OR,95%CI, P)
Past ACS exposure	Male	0.54 (0.39–0.75, P < 0.001)	0.63 (0.15–1.11, P = 0.010)	17.39 P < 0.001	0.48 (0.33–0.70, P = 0.002)
	Female	0.69 (0.48–0.99, P = 0.048)	0.42 (0.08–0.75, P = 0.015)	24.01 P = 0.48	0.94 (0.73–1.22, P = 0.644)
(1) Prior exposure to training programs	Male	0.64 (0.42–0.98, P = 0.036)	0.59 (0.13–1.26, P = 0.041)	23.63 P = 0.02	0.70 (0.45–1.08, P = 0.103)
	Female	0.87 (0.64–1.21, P = 0.417)	0.57 (0.13–1.02, P = 0.012)	37.08 P = 0.44	0.93 (0.67–1.30, P = 0.667)
(2) Ever witness ACS in families	Male	0.64 (0.45–0.90, P = 0.011)	0.58 (0.04–1.12, P = 0.035)	18.38 P = 0.04	0.58 (0.42–0.80, P < 0.001)
	Female	0.71 (0.52–0.97, P = 0.0313)	0.43 (0.05–0.80, P = 0.026)	16.75 P = 0.08	0.80 (0.60–1.06, P = 0.111)
(3) Ever personally experienced ACS	Male	0.84 (0.56–1.27, P = 0.401)	0.56 (−0.07–1.19, P = 0.083)	29.99 P = 0.32	0.91 (0.60–1.40, P = 0.674)
	Female	1.11 (0.73–1.72, P = 0.633)	0.09 (−0.47–0.65, P = 0.757)	4.4 P = 0.84	1.13 (0.74–1.75, P = 0.587)

^a Model 1 assessed the total effect of past ACS exposure (as a whole) and its three elements on decision delay, stratified by gender, with adjustments for age, education, job category, and monthly income
^b Model 2 assessed the association between past ACS exposure (as a whole), its three elements and ACS symptom knowledge, stratified by gender, with adjustments for age, education, job category, and monthly income
^c Model 3 assessed the mediation via knowledge from past ACS exposure (as a whole), its three elements to decision delay, stratified by gender, with adjustments for age, education, job category, and monthly income. The estimate of mediation effect and its statistical significance was based on 1000 times of bootstrapping

ISSN: 1471-227X