Table 3.

Incense use and cardiovascular mortality in the Singapore Chinese Health Study, stratified by baseline history of CVD, smoking, or education level at baseline (1993–2011).^a

Exposure	Participants	Person-years	CVD		CHD		Stroke
			Deaths	HR (95% CI)	Deaths	HR (95% CI)	Deaths	HR (95% CI)
Stratified by baseline CHD or strokeb
Without baseline CHD or stroke
Noncurrent use	13,774	204,917	757	1.00	421	1.00	203	1.00
urrent use	46,082	681,316	3,378	1.16 (1.07, 1.26)	1,847	1.13 (1.01, 1.26)	972	1.24 (1.06, 1.45)
With baseline CHD or stroke
Noncurrent use	863	10,447	230	1.00	142	1.00	50	1.00
urrent use	2,538	30,081	678	1.00 (0.86, 1.17)	441	1.04 (0.85, 1.27)	156	1.06 (0.76, 1.48)
Stratified by smoking statusc
Never-smokers
Noncurrent use	10,954	165,511	582	1.00	314	1.00	160	1.00
urrent use	32,976	499,294	2,144	1.12 (1.02, 1.23)	1,137	1.10 (0.96, 1.25)	672	1.26 (1.05, 1.51)
Former smokers
Noncurrent use	1,647	22,125	169	1.00	100	1.00	37	1.00
urrent use	5,346	71,726	651	1.19 (1.00, 1.42)	407	1.25 (1.00, 1.57)	154	1.18 (0.81, 1.71)
Current smokers
Noncurrent use	2,036	27,728	236	1.00	149	1.00	56	1.00
urrent use	10,298	140,377	1,261	1.05 (0.91, 1.22)	744	1.00 (0.83, 1.20)	302	1.03 (0.77, 1.38)
Stratified by education leveld
No formal education
Noncurrent use	2,333	33,063	272	1.00	151	1.00	74	1.00
urrent use	15,000	215,540	1,650	0.98 (0.86, 1.11)	884	0.93 (0.78, 1.11)	519	1.13 (0.88, 1.45)
Primary school
Noncurrent use	5,065	73,625	383	1.00	223	1.00	94	1.00
urrent use	22,985	336,371	1,844	1.14 (1.02, 1.27)	1,068	1.10 (0.95, 1.27)	484	1.30 (1.04, 1.63)
≥ Secondary school
Noncurrent use	7,239	108,676	332	1.00	189	1.00	85	1.00
urrent use	10,635	159,485	562	1.25 (1.09, 1.44)	336	1.28 (1.06, 1.53)	125	1.17 (0.88, 1.55)
aThe p for interactions between incense use (binary variable) and baseline CHD/stroke (binary variable) were 0.02, 0.15, and 0.28 for CVD, CHD, and stroke mortality, respectively. The p for interactions between incense use (binary variable) and smoking status (never, former, and current smokers) were 0.27, 0.21, and 0.21 for CVD, CHD, and stroke mortality, respectively. The p for interactions between incense use (binary variable) and education level (no formal education, primary school, and ≥ secondary school) were 0.03, 0.03, and 0.81 for CVD, CHD, and stroke mortality, respectively. bThe estimates were generated using Cox proportional hazards models, with adjustment for age at recruitment; year of recruitment; gender; dialect; education; BMI; alcohol consumption; years of smoking; amount of smoking; years since quitting smoking; moderate physical activity; duration of sleep; daily energy intake; dietary intake of vegetables, fruits, fiber, and polyunsaturated fatty acids; and self-reported history of physician-diagnosed hypertension, diabetes, and cancer. cThe estimates were generated using Cox proportional hazards models, with adjustment for age at recruitment; year of recruitment; gender; dialect; education; BMI; alcohol consumption; moderate activity; duration of sleep; daily energy intake; dietary intake of vegetables, fruits, fiber, and polyunsaturated fatty acids; and self-reported history of physician-diagnosed hypertension, diabetes, cancer, CHD, and stroke. dThe estimates were generated using Cox proportional hazards models, with adjustment for age at recruitment; year of recruitment; gender; dialect; BMI; alcohol consumption; years of smoking; amount of smoking; years since quitting smoking; moderate physical activity; duration of sleep; daily energy intake; dietary intake of vegetables, fruits, fiber, and poly­unsaturated fatty acids; and self-reported history of physician-diagnosed hypertension, diabetes, cancer, CHD, and stroke.