Although snuff use raises risk of mortality and oral cancer incidence globally «Siddiqi K, Shah S, Abbas SM, ym. Global burden of ...»1, «Asthana S, Labani S, Kailash U, ym. Association of...»3, «Hansson J, Galanti MR, Hergens MP, ym. Snus (Swedi...»4, «Sinha DN, Suliankatchi RA, Gupta PC, ym. Global bu...»5 the Swedish snuff (which is the type of snuff used in Finland) users have only small increases in risk of death and no increases in oral cancer risk in prospective observational studies conducted in Sweden «Hansson J, Galanti MR, Hergens MP, ym. Snus (Swedi...»4, «Byhamre ML, Araghi M, Alfredsson L, ym. Swedish sn...»6, «Araghi M, Galanti MR, Lundberg M, ym. No associati...»7. However, it should be noted that the Swedish snuff studies focus on males. Higher mortality could be explained by increased cardiovascular deaths «Byhamre ML, Araghi M, Alfredsson L, ym. Swedish sn...»6.
Description of the included studies
Reference | Study type | Population | Exposure and comparison | Outcomes | Risk of bias «Additional comments for included studies»1 | |
---|---|---|---|---|---|---|
«Siddiqi K, Shah S, Abbas SM, ym. Global burden of ...»1 | SR/MA | 32 case/control and cohort studies included in the meta-analysis. Studies grouped geographically and areas analyzed separately in a meta-analysis: India, Pakistan, Sweden, Norway, North America Inclusion criteria: smoking and alcohol use adequately controlled. Exclusion criteria: cross-sectional studies, case series and case reports. | Smokeless tobacco (SLT) use vs. population based risk | Risk for oral, pharyngeal, oesophaegal cancers or ischaemic heart disease Burden of the SLT use expressed as proportion of the disability-adjusted life-years (DALYs) lost and deaths reported |
No risk of bias assessment of the included studies. | |
«Wyss AB, Hashibe M, Lee YA, ym. Smokeless Tobacco ...»2 | Pooled analysis of studies | 11 US case-control studies (1981–2006) included in the pooled analysis (6772 cases and 8375 controls) Approx. half of the cases and controls were between 50 and 65 years of age, majority were non-Hispanic white, over two-thirds were male. Individuals excluded if: missing data on frequency of alcohol drinking or duration of cigarette, cigar, or pipe smoking SLT product use not separately described | SLT use: snuff and chewing tobacco separately analyzed for never and ever cigarette smokers | Head and neck cancer: oral, gum, pharynx, hypopharynx, oropharynx or larynx, all sites, also male and female separately | Not reported | |
«Asthana S, Labani S, Kailash U, ym. Association of...»3 | SR/MA | 37 case-control or cohort studies (1960–2016) from Southeast Asia region (SEAR), the Eastern Mediterranean Region (EMR), Europe, and region of Americas (North and South) Exposure status in the studies included was ever versus never SLT use and should be adjusted for at least smoking status. All age groups 10-99 y included, in most studies mean age around 50 years | SLT use among never versus ever users | Oral cancer risk in different regions, concerning different SLT products and both sexes | Not reported | |
«Hansson J, Galanti MR, Hergens MP, ym. Snus (Swedi...»4 | Pooled analysis of studies | 8 Swedish cohort studies (1978-2004) 130 485 men who never smoked | Current snuff users compared to noncurrent snuff users (adjusted to age and BMI) | Incidence and survival after stroke | ||
«Sinha DN, Suliankatchi RA, Gupta PC, ym. Global bu...»5 | SR/MA | 16 studies worldwide | SLT use compared to non-tobacco use | All-cause mortality (OR) and subgroup analysis for different mortality causes and different regions | Bias on the included studies mainly: Confounding Exposure and Selection also significant heterogeneity between studies | |
«Byhamre ML, Araghi M, Alfredsson L, ym. Swedish sn...»6 | Pooled analysis | 8 prospective cohort studies from Sweden (1978–2010), including 169 103 never smoking men | Current snuff users compared to never tobacco users (including snuff use) | All-cause mortality (OR) and subgroup analysis for different mortality causes | Not reported for individual studies. Check comments below. |
|
«Araghi M, Galanti MR, Lundberg M, ym. No associati...»7 | Pooled analysis | 9 prospective cohort studies from Sweden including 418,369 male participants, 9,201,647 person years, from 1978 to 2013 | Never snuff users to ever snuff users or current snuff users | Oral cancer | Not reported for individual studies. Check comments below. |
|
RCT = randomized controlled trial; SR = systematic review; MA = meta-analysis RCT = randomized controlled trial; SR = systematic review; MA = meta-analysis |
Results
Reference | Number of studies and number of patients | Follow-up time | Absolute risk I or E | Absolute risk C | Relative effect (95 % CI) |
---|---|---|---|---|---|
«Siddiqi K, Shah S, Abbas SM, ym. Global burden of ...»1 | 32 studies | - | - | - | SLT use vs. population risk OR India 5,12 (3,27–8,02) Pakistan 8,81 (3,14–24,69) Sweden 0,92 (0,68–1,25) Norway 1,10 (0,50–2,42) North America 0,95 (0,70–1,28) |
«Wyss AB, Hashibe M, Lee YA, ym. Smokeless Tobacco ...»2 | pooled data of 11 US case-control studies 6772 cases 8375 controls | - | 20/82 10/72 20/32 24/74 |
379/3435 52/3108 391/1735 737/2449 |
Never cigarette smokers: never snuff use vs. ever snuff use, cancer risk (OR) in oral cavity 3,01 (1,63–5,55) gum 12,70 (4,76–33,7) female 8,89 (3,59–22,0) male 0,86 (0,49–1,51) |
«Asthana S, Labani S, Kailash U, ym. Association of...»3 | 37 studies globally | - | - | - | Ever snuff use vs. Never snuff use OR (adjusted for smoking status) South East Asia 4,44 (3,51–5,61) East mediterranean 1,28 (1,04–1,56) European 0,86 (0,58–1,29) American 4,72 (0,66–33,62) Significantly higher risk was found for females 5,83 (2,93–11,58) Different products: gutkha 8,67 (3,59–20,93) pan tobacco / betel liquid 7,18 (5,48–9,41) oral snuff 4,18 (2,37–7,38) Mainpuri tobacco 3,32 (1,32–8,36) Swedish snuff 0,86 (0,58–1,29) |
«Araghi M, Galanti MR, Lundberg M, ym. No associati...»7 | 9 studies, 418,369 male participants, 9,201,647 person years | 5-35 years | Ever-users vs. never-users HR 0,9; CI 0,74-1,09. Current users vs. never-users HR 0,79; CI 0,63-1,00. Adjusted for Age, smoking and BMI. |
||
Level of evidence: Low | |||||
I = intervention; E = exposure; C = comparison; CI = confidence interval |
Studies with combined risk for oral and oropharyngeal or head and neck cancers are not included in the table Outcome: oral cancer
Reference | Number of studies and number of patients (E/C) | Follow-up time | Absolute risk E | Absolute risk C | Relative risk (95 % CI) |
---|---|---|---|---|---|
«Siddiqi K, Shah S, Abbas SM, ym. Global burden of ...»1 | 32 studies | Worldwide 6,4 milj DALYs lost and 266 592 deaths /year | |||
«Hansson J, Galanti MR, Hergens MP, ym. Snus (Swedi...»4 | pooled analysis | mean follow-up time 17 years | - | - | Current snuff user HR for stroke 1,04 (95% CI 0,92–1,17). Mean age at stroke diagnosis was 62 and 68 years amongst current and noncurrent snuff users, respectively (P < 0,0001). HR for death during the follow-up period 1,32 (95% CI 1,08–1,61). |
«Sinha DN, Suliankatchi RA, Gupta PC, ym. Global bu...»5 | 16 studies | - | - | - | All-cause mortality (1,22; 1,11–1,34), all cancer (1,31; 1,16–1,47),
UADT cancer (2,17; 1,47–3,22), stomach cancer (1,33; 1,12–1,59), cervical
cancer (2,07; 1,64–2,61), IHD (1,10; 1,04–1,17) and stroke (1,37; 1,24–151). Major regional differences in subgroup analysis |
«Byhamre ML, Araghi M, Alfredsson L, ym. Swedish sn...»6 | 8 studies, 169 103 participants | 10 years | All-cause mortality 1,28; CI 1,20-1,35. Cardiovascular disease mortality 1,27; CI 1,15-1,41. Cancer mortality 1,12; CI 1,00-1,26. All estimates are adjusted for age and BMI. | ||
Level of evidence: Low | |||||
E = exposure; C = comparison; CI = confidence interval |
Comment
Pooled studies about Swedish snuff use have been mainly conducted with self-administered questionnaires, which could lead to selection bias. Although heterogeneity of results between Swedish studies have not been reported, several sensitivity analyses were conducted regarding the mortality «Byhamre ML, Araghi M, Alfredsson L, ym. Swedish sn...»6. There was a statistically significant dose-response regarding the duration of snuff use but not with the amount used per week «Araghi M, Galanti MR, Lundberg M, ym. No associati...»7. As the effect sizes are small for mortality and the evidence is observational, strong judgements about the causality can't be made.