Review with meta-analysis relating North American, European and Japanese snus or smokeless tobacco use to major smoking-related diseases

Table 1 Studies considered in analyses of smokeless tobacco risk among never smokers in the United States

Study	Ref.	Study type1	Timing	Population	Diseases considered	Excludes pipe/cigar	Exposure index
Main sources
CPS-I2	Henley et al[9], 2005	P	1959 to 1971	Families of volunteers’ friends and neighbours	LC, COPD3, IHD, Stroke	Yes	Current vs never
CPS-II2	Henley et al[9], 2005	P	1982 to 2000	Families of volunteers’ friends and neighbours	LC, COPD4, IHD, Stroke	Yes	Current vs never
NHIS5	Inoue-Choi et al[10], 2019; Inoue-Choi et al[11], 2020	P	1991-2010 to 2015	Civilian non-institutionalized	LC, COPD6, IHD, Stroke	Yes	Current vs never
NLMS7	Timberlake et al[13], 2017	P	1985-2011 to 2011	Civilian non-institutionalized	IHD, Stroke	Yes	Current vs never
NLMS7	Fisher et al[8], 2019	P	1993-2005 to 2010	Civilian non-institutionalized	LC	Yes	Current vs never
Other sources
NHANES8	Accortt et al[6], 2002	P	1971-75 to 1992	Civilian non-institutionalized	LC, IHD, Stroke	No	Ever vs never
AHS9	Andreotti et al[7], 2017	P	1993-97 to 2010-11	Pesticide applicators and their spouses	LC	Yes	Ever vs never

¹Prospective study.

²CPS: Cancer Prevention Study.

³Respiratory symptom diseases (ICD7 470-527).

⁴Chronic obstructive pulmonary disease (ICD9 490-492, 496).

⁵NHIS: National Health Interview Surveys.

⁶Chronic lower respiratory disease (ICD10 J40-J47).

⁷NLMS: National Longitudinal Mortality Study.

⁸NHANES1: First National Health and Nutrition Examination Survey.

⁹AHS: Agricultural Health Study.

COPD: Chronic obstructive pulmonary disease.

Table 2 Relative risks in analyses of smokeless tobacco risk among never smokers in the United States

Study	Sex	Lung cancer	Chronic obstructive pulmonary disease	Ischaemic heart disease	Stroke	Adjustment factors
Main sources
CPS-I	M	1.08 (0.64-1.83)	1.86 (1.12-3.06)	1.12 (1.03-1.21)	1.46 (1.31-1.64)	Age, alc, asp, bmi, edu, ex, fat, f/v, race
CPS-II	M	2.00 (1.23-3.24)	1.28 (0.71-2.32)	1.26 (1.08-1.47)	1.40 (1.10-1.79)	Age, alc, asp, bmi, edu, emp, ex, fat, f/v, race
NHIS	M + F	1.43 (0.51-4.01)	1.35 (0.39-4.76)	1.66 (1.30-2.13)	1.09 (0.56-2.11)	Age, edu, race, sex, year
NLMS	M + F	2.98 (0.91-9.76)	-	1.24 (1.05-1.46)	0.92 (0.67-1.27)	Lung cancer: age, edu, hea, inc, race, sexIHD and CVD: age, edu, inc, race, sex
Random-effects meta-analysis		1.59 (1.06-2.39) (n = 4)	1.57 (1.09-2.26) (n = 3)	1.26 (1.10-1.45) (n = 4)	1.27 (1.03-1.57) (n = 4)
Other sources
NHANES	M	-	-	0.6 (0.3-1.2)	0.7 (0.2-2.0)	Lung cancer: age, alc, ex, f/v, pov, race, regIHD: age, alc, bmi, chol, ex, f/v, pov, race, sbpCVD: age, alc, ex, f/v, pov, race, sbp
NHANES	F	9.1 (1.1-75.4)	-	1.4 (0.8-2.2)	1.0 (0.3-2.9)
AHS	M + F	2.21 (1.11-4.42)	-	-	-	Age, alc, edu, race, reg, sex
All sources
Random-effects meta-analysis		1.80 (1.23-2.64) (n = 6)	1.57 (1.09-2.26) (n = 3)	1.24 (1.08-1.43) (n = 6)	1.24 (1.02-1.52) (n = 6)

Alc: Alcohol, asp: Aspirin use; bmi: Body mass index; chol: Cholesterol; edu: Education; emp: Employment, ex: Exercise; f/v: Fruit and vegetable intake; hea: Health status; inc: Income; pov: Poverty; reg: Region; sbp: Systolic blood pressure; year: Year of survey.

Table 3 Studies considered in analysis of current snus use among non-smokers in Scandinavia

Study1	Source	Study type2	Timing	Population	Diseases considered	Unexposed snus3
CWC	Bolinder et al[17], 1994	P	1971-74 to 1985	Construction workers	LC, IHD, stroke	Non
	Hergens et al[23], 2007		1978-93 to 2004		AMI	Never
	Luo et al[29], 2007		1978-92 to 2004		LC	Never
	Hergens et al[24], 2008		1978-92 to 2003		CVD	Never
	Hansson et al[20], 2012; Hansson et al[21], 2014		1978-93 to 2004		AMI, stroke	Non
MALMÖ	Janzon and Hedblad[27], 2009	P	1991-96 to 2004	Population-based, Malmö city	AMI, stroke	Non
	Hansson et al[20], 2012; Hansson et al[21], 2014				AMI, stroke	Non
MONICA	Asplund et al[15], 2003	NCC	1986-99 to 2000	Population-based, Norrbotten and Västerbotten counties	CVD	Non
	Wennberg et al[31], 2007	NCC	1986-99 to 1999		AMI	Never
	Huhtasaari et al[25], 1992	CC	1989-91		AMI	Non
	Huhtasaari et al[26], 1999	CC	1991-93		AMI	Non
	Hansson et al[20], 2012; Hansson et al[21], 2014	P	1986-2004 to 2004		AMI, stroke	Non
NMC	Hansson et al[20], 2012; Hansson et al[21], 2014	P	1997 to 2004	Participant in charity walk	AMI, stroke	Non
NORWAY	Boffetta et al[16], 2005	P	1964-67 to 2001	Population sample and relatives of emigrants	LC	Ever vs never
SALLS	Johansson et al[28], 2005	P	1988-89 to 2000	Civilian non-institutionalized	IHD	Non
SALT	Hansson et al[19], 2009	P	1998-2002 to 2005	Twins born in Sweden 1926-1958	IHD, stroke	Never
	Hansson et al[20], 2012; Hansson et al[21], 2014		1998-2002 to 2004		AMI, stroke	Non
Scania-PHC	Hansson et al[20], 2012; Hansson et al[21], 2014	P	2002 to 2004	Population-based, Skåne County	AMI, stroke	Non
Stockholm-PHC	Hansson et al[20], 2012; Hansson et al[21], 2014	P	2002 to 2004	Population-based, Stockholm County	AMI, stroke	Non
ULF	Haglund et al[18], 2007	P	1988-89 to 2003	Civilian, non-institutionalized	IHD, stroke	Non
Two Counties	Hergens et al[22], 2005	CC	1992-94	Randomly selected, Stockholm and Västernorrland counties	AMI	Never
WOLF	Hansson et al[20], 2012; Hansson et al[21], 2014	P	1992-98 to 2004	Employed in three counties	AMI, stroke	Non

¹CWC: Construction workers cohort; MONICA: Monitoring of trends in cardiovascular disease; NMC: National March Cohort; PHC: Public Health Cohort; SALLS: Swedish Annual Level of Living Survey; SALT: Screening across the lifespan twin study; ULF: Swedish survey of living conditions; WOLF: Work, lipids and fibrinogen.

²CC: Case control; NCC: Nested case control, P: Prospective.

³Exposed group: Current unless stated. In some studies the unexposed group may include non-regular tobacco users.

Table 4 Relative risks in analyses of ischaemic heart disease/acute myocardial infarction and stroke in relation to current snus use among never smokers in Scandinavia

Study	Source1	Current vs never		Current vs non		Adjustmentfactors2
		IHD/AMI	Stroke	IHD/AMI	Stroke
CWC	Bolinder et al[17], 1994	-	-	1.35 (1.13-1.62)3	1.29 (0.83-1.99)3	Age, res
CWC	Hergens et al[23], 2007	1.02 (0.92-1.14)	-	1.03 (0.93-1.15)	-	Age, BMI, res
CWC	Hergens et al[24], 2008	-	1.05 (0.95-1.17)	-	1.06 (0.96-1.18)	Age, BMI, res
CWC	Hansson et al[20], 2012; Hansson et al[21], 2014	-	-	1.01 (0.90-1.14)	1.03 (0.90-1.17)	Age, BMI4
MALMÖ	Janzon and Hedblad[27], 2009	-	-	0.75 (0.30-1.80)	0.59 (0.20-1.50)	Age, BMI, dia, hyp, mar, occ, phys
MALMÖ	Hansson et al[20], 2012; Hansson et al[21], 2014	-	-	1.00 (0.37-2.70)	1.23 (0.50-2.99)	Age, BMI4
MONICA	Asplund et al[15], 2003	-	-	-	0.87 (0.41-1.83)	Age, chol, cohort, edu, dia, hyp, mar, year
MONICA	Wennberg et al[31], 2007	0.82 (0.46-1.43)	-	0.85 (0.48-1.50)	-	Age, BMI, chol, edu, phys, res, year
MONICA	Huhtasaari et al[25], 1992	-	-	0.89 (0.62-1.29)	-	Age
MONICA	Huhtasaari et al[26], 1999	-	-	0.58 (0.35-0.94)	-	Age, chol, dia, edu, her, hyp, mar, res
MONICA	Hansson et al[20], 2012; Hansson et al[21], 2014	-	-	0.77 (0.35-1.69)	0.65 (0.23-1.80)	Age, BMI4
NMC		-	-	No IHD cases in current snus users	1.28 (0.40-4.10)	Age, BMI4
SALLS		1.41 (0.61-3.28)	-	-	-	Age, BMI, dia, hyp, phys
SALT	Hansson et al[19], 2009	0.85 (0.51-1.41)	1.18 (0.67-2.08)	0.85 (0.51-1.40)	1.15 (0.66-2.02)	Age, chol, dia, hyp
SALT	Hansson et al[20], 2012; Hansson et al[21], 2014	-	-	1.56 (0.98-2.48)	0.98 (0.52-1.83)	Age, BMI4
Scania-PHC		-	-	1.90 (0.90-4.00)	3.17 (1.50-6.70)	Age, BMI4
Stockholm-PHC		-	-	1.21 (0.48-3.08)	0.58 (0.14-2.45)	Age, BMI4
ULF		-	-	1.15 (0.54-2.41)	1.01 (0.35-2.92)	Age, heal, ill, phys, res, ses
Two Counties		0.73 (0.35-1.50)	-	0.73 (0.35-1.51)	-	Age, area
WOLF		-	-	3.30 (0.63-17.1)	0.96 (0.28-3.30)	Age, BMI4

¹See Table 2 for source if the study is only analysed by one publication or by the two pooled analyses by Hansson et al[20] only.

²Abbreviations used: BMI: Body mass index; chol: Cholesterol; dia: Diabetes; edu: Education; heal: Self-reported health; her: Heredity; hyp: Hypertension; ill: Self-reported longstanding illnesses; mar: Marital status; occ: Occupation; phys: Physical activity; res: Region of residence; ses: Socioeconomic status; year: Recruitment year.

³Estimated from results given for two groups by age at entry to the study.

⁴Body mass index adjusted for in the analyses of stroke, but not acute myocardial infarction.

All results are for men. Where results in any row are given for both comparison groups (never and non) for the same disease, the result for the comparison group non were estimated from data provided in the source paper.

Table 5 Meta-analyses of ischaemic heart disease/acute myocardial infarction and stroke results in relation to snus use among never smokers in Scandinavia

Disease	Comparison group	All data or latest	Random-effects meta-analysis relative risk (95%CI)	Heterogeneity
				Chi sq	DF	P value
IHD/AMI	Never	Latest	1.00 (0.91-1.11)	2.33	4	NS
	Non	All	1.04 (0.92-1.18)	24.87	15	0.052
		Latest	1.10 (0.92-1.33)	9.18	8	NS
Stroke	Never	Latest	1.05 (0.95-1.17)	0.16	1	NS
	Non	All	1.06 (0.98-1.14)	12.69	13	NS
		Latest	1.12 (0.86-1.45)	10.26	8	NS

Where the comparison group is non users, there are (see Table 4) estimates for some studies from multiple publications. For these studies, the estimate “Latest” includes only the result from the latest publication, while the estimate “All” includes all the results. Where the comparison group is never users, no study provides more than one estimate. NS: Not significant (P ≥ 0.1).