Table 2. Pooled analysis for stomach cancer mortality and incidence by exposure lag (latency) and type of outcome using random effects model.

	Outcome	na	SMR	95% CI	τ2 b	PQc	I2 (%)d
Men
All exposure lags	I	14	1.09	(0.94–1.26)	0.019	0.16	28.0
	M	30	1.16	(1.00–1.34)	0.085	<0.001	63.5
	M+I	44	1.13	(1.02–1.26)	0.057	<0.001	54.7
At least 10 yr exposure lag	I	2	–	–	–	–	–
	M	9	1.16	(0.79–1.69)	0.213	<0.001	75.3
	M+I	11	1.09	(0.77–1.53)	0.197	<0.001	71.8
No exposure lag	I	14	1.09	(0.94–1.26)	0.019	0.16	28.0
	M	26	1.18	(0.99–1.40)	0.111	<0.001	69.6
	M+I	40	1.14	(1.01–1.28)	0.069	<0.001	59.7
Women
All exposure lags	I	6	1.1	(0.52–2.33)	0	0.99	–
	M	13	0.93	(0.67–1.30)	0	0.90	0
	M+I	19	0.96	(0.71–1.30)	0	0.99	–
No exposure lag	I	6	1.1	(0.52–2.33)	0	0.99	–
	M	12	0.89	(0.62–1.26)	0	0.90	0
	M+I	18	0.92	(0.67–1.27)	0	0.99	–
Total cohort
All exposure lags	I	15	1.07	(0.91–1.25)	0.022	0.26	25.5
	M	40	1.17	(1.03–1.33)	0.087	<0.001	69.1
	M+I	55	1.15	(1.03–1.27)	0.069	<0.001	61.9
At least 10 yr exposure lag	I	0	–	–	–	–	–
	M	10	1.12	(0.80–1.56)	0.182	<0.001	81.6
	M+I	12	1.07	(0.79–1.44)	0.169	<0.001	78.5
No exposure lag	I	15	1.07	(0.91–1.25)	0.022	0.26	25.5
	M	36	1.18	(1.03–1.36)	0.102	<0.001	72.7
	M+I	51	1.15	(1.03–1.28)	0.078	<0.001	64.8

Abbreviations: – = not applicable; CI=confidence interval; I=incidence, M=mortality.

No results for women for at least 10 year exposure lag as only one mortality study reported a SMR.

^aNumber of cohorts included.

^bVariance (amount of heterogeneity).

^cP-value for the heterogeneity test.

^dPercentage of total variability due to heterogeneity.