Systematic Reviews
Copyright ©The Author(s) 2020.
World J Meta-Anal. Jun 28, 2020; 8(3): 245-264
Published online Jun 28, 2020. doi: 10.13105/wjma.v8.i3.245
Table 1 Studies in meta-analysis of lung cancer and exclusive cigar smoking1
Study name2Data sourceStudy typeYears of follow-up3SexAge groupStudy size4
BOUCOTBoucot et al[20], 1972P10M45+6027
CPS IHammond[21], 1966P4M35-844405585
Hammond[22], 1972P6M35-84
Shanks et al[23], 1998P13M35+
CPS IIJacobs et al[24], 1999P12M30+5085766
Shapiro et al[25], 2000P12M30+
DORN (US veterans study)Dorn[26], 1959P2M30+2480467
Kahn[27], 1966P8M35-84
Rogot et al[28], 1980P16M31-84
HAMMONHammond et al[18,19], 1958P3M50-69187783
KAISERIribarren et al[13], 1999P25M30-8517774
LEVINLevin[29], 1954C-M35+2855
MALHOTMalhotra et al[14], 2017P15M55-625244408
NLMSChristensen et al[17], 2018P26C35-80146529
SADOWSSadowsky et al[30], 1953C-MAll2605
WYNDE7Higgins et al[31], 1988C-M20-8060339
1Compared with never smoking any tobacco product.
2Study name is an identifier assigned by the authors based on the published study name or the name of the first author of a source paper. This identifier is used in the meta-analysis output. Note that studies KAISER and MALHOT were included in sensitivity analysis only.
3Years of follow-up is relevant only to prospective studies.
4Study size: Total number of participants in the study, irrespective of smoking habits but taking account of sex.
5Size not given for second source, and a slightly different number of 442455 given for third source.
6Slightly different size of 508353 given for second source.
7Size taken from third source, other sources only gives person-years. The numbers include about 0.5% females.
8Includes 74770 from two non-United States studies.
9Size is for sexes combined. P: Prospective; C: Case control; M: Male; C: Combined male and female.
Table 2 Studies in meta-analysis of chronic obstructive pulmonary disease and exclusive cigar smoking1
Study name2Data sourceStudy typeYears of follow-upSexAge groupStudy size3
CPS IShanks et al[23], 1998P13M35+442455
DORN (US veterans study)Kahn[27], 1966P8M35-842480464
KAISERIribarren et al[13], 1999P25M30-8517774
NLMSChristensen et al[17], 2018P26C35-80146529
1Compared with never smoking any tobacco product.
2Study name is an identifier assigned by the authors based on the published study name or the name of the first author of the source paper. This identifier is used in the meta-analysis output. Note that study KAISER was included in sensitivity analysis only.
3Study size: Total number of participants in the study, irrespective of smoking habits but taking account of sex.
4See Table 1 for study size. P: Prospective; M: Male; C: Combined male and female.
Table 3 Studies on ischaemic heart disease, stroke and circulatory disease and exclusive cigar smoking1
Study name2Data sourceStudy typeYears of follow-up3SexAge groupStudy size4Disease data available and notes5
CPS IHammond[21], 1966P4M35-844405586IHD: CHD (ICD-6 420); Stroke: -; Circ: -
Shanks et al[23], 1998P13M35+IHD: CHD; Stroke: CVD; Circ: -
CPS IIJacobs et al[32], 1999P9M30+508576IHD: CHD; Stroke: -; Circ: -
DORN (US veterans study)Dorn[26], 1959P2M30+2480466IHD: -; Stroke: -; Circ: Cardiovascular diseases (ICD-6 330-334, 400-468)
Kahn[27], 1966P8M35-84IHD: Arteriosclerotic (coronary) heart disease (ICD-6 420); Stroke: Cerebral vascular lesions (ICD-6 330-334); Circ: Total cardiovascular disease (ICD-6 330-334, 400-468)
Rogot et al[28], 1980P16M31-84IHD: CHD (ICD-6 420); Stroke: Stroke (ICD-6 330-334); Circ: Cardiovascular (ICD-6 330-334, 400-468)
HAMMONHammond et al[19], 1958P3M50-69187783IHD: Coronary artery disease; Stroke: Cerebral vascular diseases; Circ: Heart and circulatory diseases (ICD-6 330-334, 400-468)
KAISERIribarren et al[13], 1999P25M30-8517774IHD: CHD; Stroke: Ischaemic stroke and haemorrhagic stroke combined; Circ: -
KAUFMAKaufman et al[33], 1987C-M40-5415067IHD: Non-fatal myocardial infarction; Stroke: -; Circ: -
NHISInoue-Choi et al[16], 2019P17C18-9571314IHD: - (“Coronary heart disease” used too broad a definition); Stroke: Cardiovascular disease (ICD-10 I60-I69); Circ: Combined “Coronary heart disease” and Stroke; this excludes some hypertensive disease, diseases of arteries and diseases of veins and other circulatory disease – included in sensitivity analysis only
Rostron et al[15], 20198X-C35+Not specifiedIHD: Heart conditions; this includes angina, CHD, heart attack and other heart disease – included in sensitivity analysis only; Stroke: Stroke; Circ: -
NLMSChristensen et al[17], 2018P26C35-80146529IHD: - (“Cardiovascular diseases” used too broad a definition); Stroke: Cerebrovascular disease (ICD-10 I60-I69); Circ: Circulatory disease (ICD-10 I00-I09, I20-I25, I26-I28, I29-I51, I60-I69, I70, I71, I72-I78); this definition excludes “Diseases of veins” and “Other circulatory diseases” (ICD-10 I80-I89) – included in sensitivity analysis only
1Compared with never smoking any tobacco product.
2Study name is an identifier assigned by the authors based on the published study name or the name of the first author of the source paper. This identifier is used in the meta-analysis output. Note that some study results were included in sensitivity analysis only: The results from study KAISER (each disease); study NHIS (the prospective result for circulatory disease and the cross-sectional result for IHD); and study NLMS (the result for circulatory disease).
3Years of follow-up is relevant only to prospective studies.
4Study size: Total number of participants in the study, irrespective of smoking habits but taking account of sex.
5The diseases considered were IHD, stroke and circ. The text gives, for each of these diseases, the disease definition as given in the paper together with additional notes where relevant (reason for rejection, method of derivation, problems with the result that mean it should be included in the sensitivity analysis only). The ICD codes relating to IHD, stroke and circ can be found at https://coder.aapc.com/icd-10-codes-range/110.
6See Table 1 for study size.
7Study limited to non-cigarette smokers.
8This is a cross-sectional analysis of baseline data: Results from the prospective study (above) were used in the meta-analysis when possible. IHD: Ischaemic heart disease; Circ: Stroke and circulatory disease P: Prospective; C: Case control; X: Cross-sectional; M: Male; C: Combined male and female; ICD: The International Classification of Disease.
Table 4 Individual estimates used in the meta-analyses1
DiseaseStudy name2SourceExposureRelative risk (95%CI)
Lung cancerBOUCOT[20]Ever8.81 (0.45 to 170.58)
CPSI[23]Current3.30 (2.68 to 4.06)
[22]Ever2.11 (1.45 to 3.07)
CPSII[25]Current5.10 (4.00 to 6.60)
[25]Former1.60 (1.20 to 2.40)
[25]Ever3.01 (2.42 to 3.74)
DORN[28]Current1.66 (1.18 to 2.34)
[27]Former1.02 (0.41 to 2.54)
[27]Ever1.49 (0.97 to 2.27)
HAMMON[19]Ever1.02 (0.42 to 2.51)
KAISER[13]Current32.14 (1.12 to 4.11)
LEVIN[29]Ever1.41 (0.76 to 2.60)
MALHOT[14]Ever32.73 (2.06 to 3.60)
NLMS[17]Current3.26 (1.86 to 5.71)
[17]Former1.35 (0.70 to 2.61)
[17]Ever2.04 (1.33 to 3.15)
SADOWS[30]Ever2.98 (1.06 to 8.33)
WYNDE7[31]Current3.15 (1.78 to 5.57)
[31]Former2.46 (1.27 to 4.77)
[31]Ever2.83 (1.78 to 4.51)
COPDCPSI[23]Current1.42 (0.96 to 2.03)
DORN[27]Current0.79 (0.31 to 2.03)
[27]Former1.38 (0.42 to 4.51)
[27]Ever0.94 (0.43 to 2.04)
KAISER[13]Current31.45 (1.10 to 1.91)
NLMS[17]Current2.44 (0.98 to 6.05)
[17]Former0.05 (0.00 to 3.19)
[17]Ever0.77 (0.32 to 1.87)
IHDCPSI[23]Current1.05 (1.00 to 1.11)
CPSII[32]Current1.13 (0.96 to 1.34)
[32]Former1.07 (0.93 to 1.24)
[32]Ever1.09 (0.98 to 1.22)
DORN[28]Current1.12 (1.05 to 1.18)
[27]Former1.41 (1.25 to 1.60)
[27]Ever1.13 (1.05 to 1.22)
HAMMON[19]Ever1.28 (1.13 to 1.44)
KAISER[13]Current31.27 (1.12 to 1.45)
KAUFMA[33]Current1.25 (0.58 to 2.67)
NHIS[15]Current30.88 (0.61 to 1.27)
[15]Former31.33 (1.03 to 1.72)
[15]Ever31.15 (0.93 to 1.41)
StrokeCPSI[23]Current0.96 (0.87 to 1.06)
DORN[28]Current1.07 (0.95 to 1.21)
[27]Former1.14 (0.85 to 1.53)
[27]Ever1.09 (0.93 to 1.29)
HAMMON[19]Ever1.33 (1.04 to 1.70)
KAISER[13]Current31.08 (0.88 to 1.33)
NHIS[16]Current1.60 (0.72 to 3.57)
[16]Former0.56 (0.20 to 1.57)
[16]Ever0.94 (0.50 to 1.77)
NLMS[17]Current0.50 (0.21 to 1.22)
[17]Former1.08 (0.66 to 1.75)
[17]Ever0.85 (0.55 to 1.30)
Circulatory diseaseDORN[28]Current1.10 (1.05 to 1.16)
[27]Former1.37 (1.24 to 1.52)
[27]Ever1.13 (1.06 to 1.20)
HAMMON[19]Ever1.27 (1.15 to 1.40)
NHIS[16]Current31.34 (0.88 to 2.06)
[16]Former30.75 (0.52 to 1.10)
[16]Ever30.93 (0.70 to 1.24)
NLMS[17]Current31.11 (0.87 to 1.42)
[17]Former31.13 (0.93 to 1.38)
[17]Ever31.12 (0.96 to 1.31)
1Compared with never smoking any tobacco product.
2Study name is an identifier assigned by the authors based on the published study name or the name of the first author of the source paper. This identifier is used in the meta-analysis output.
3These results were only included in the sensitivity analyses. COPD: Chronic obstructive pulmonary disease; IHD: Ischaemic heart disease.
Table 5 Meta-analysis results for lung cancer, chronic obstructive pulmonary disease, ischaemic heart disease, stroke and circulatory disease for cigar smoking1
DiseaseExposureNumber of studiesRelative risk (95%CI) fixed effectRelative risk (95%CI) random effectsHeterogeneity I2(2)
Lung cancerCurrent cigar smokers, excluding KAISER53.36 (2.93 to 3.84)3.12 (2.11 to 4.62)85.22 (P < 0.001)
Current cigar smokers, including KAISER63.29 (2.88 to 3.76)2.98 (2.08 to 4.26)82.65 (P < 0.001)
Former cigar smokers41.61 (1.23 to 2.09)1.61 (1.23 to 2.09)0.00 (NS)
Ever cigar smokers, excluding MALHOT92.35 (2.04 to 2.71)2.11 (1.64 to 2.72)54.77 (P < 0.05)
Ever cigar smokers, including MALHOT102.43 (2.13 to 2.75)2.22 (1.79 to 2.74)51.51 (P < 0.05)
COPDCurrent cigar smokers, excluding KAISER31.42 (1.02 to 1.96)1.42 (0.89 to 2.26)29.95 (NS)
Current cigar smokers, including KAISER41.44 (1.16 to 1.77)1.44 (1.16 to 1.77)0.00 (NS)
Former cigar smokers21.06 (0.34 to 3.31)0.47 (0.02 to 9.88)58.19 (NS)
Ever cigar smokers20.86 (0.48 to 1.54)0.86 (0.48 to 1.54)0.00 (NS)
IHDCurrent cigar smokers, excluding NHIS and KAISER41.08 (1.04 to 1.13)1.08 (1.04 to 1.13)0.28 (NS)
Current cigar smokers, including NHIS and KAISER61.09 (1.06 to 1.14)1.11 (1.04 to 1.19)48.68 (P < 0.1)
Former cigar smokers, excluding NHIS21.25 (1.14 to 1.38)1.23 (0.94 to 1.61)87.72 (P < 0.01)
Former cigar smokers, including NHIS31.26 (1.16 to 1.38)1.26 (1.03 to 1.53)75.96 (P < 0.05)
Ever cigar smokers, excluding NHIS31.15 (1.09 to 1.21)1.16 (1.06 to 1.26)51.96 (NS)
Ever cigar smokers, including NHIS41.15 (1.09 to 1.21)1.15 (1.08 to 1.23)27.95 (NS)
StrokeCurrent cigar smokers, excluding KAISER41.00 (0.93 to 1.08)1.01 (0.87 to 1.16)46.09 (NS)
Current cigar smokers, including KAISER51.01 (0.94 to 1.09)1.02 (0.92 to 1.13)33.51 (NS)
Former cigar smokers31.08 (0.85 to 1.38)1.08 (0.85 to 1.38)0.00 (NS)
Ever cigar smokers41.12 (0.98 to 1.27)1.11 (0.95 to 1.31)21.80 (NS)
Circulatory diseaseCurrent cigar smokers, excluding NHIS and NLMS11.10 (1.05 to 1.16)--
Current cigar smokers, including NHIS and NLMS31.10 (1.05 to 1.16)1.10 (1.05 to 1.16)0.00 (NS)
Former cigar smokers, excluding NHIS and NLMS11.37 (1.24 to 1.52)--
Former cigar smokers, including NHIS and NLMS31.28 (1.17 to 1.39)1.11 (0.84 to 1.46)81.92 (P < 0.01)
Ever cigar smokers, excluding NHIS and NLMS21.17 (1.11 to 1.23)1.19 (1.06 to 1.33)72.63 (P < 0.1)
Ever cigar smokers, including NHIS and NLMS41.15 (1.10 to 1.21)1.15 (1.06 to 1.26)51.29 (NS)
1Compared with never smoking any tobacco product.
2Heterogeneity between the results included in the meta-analysis, given as the I2 statistic together with its significance coded as: Not significant; P < 0.1; P < 0.05; P < 0.01; P < 0.001. The I2 statistic is calculated as 100 × (H-df)/H, where H is the heterogeneity chi-squared statistic, df is the degrees of freedom, and if the resulting value is negative I2 is taken to be 0. NS: Not significant; COPD: Chronic obstructive pulmonary disease; IHD: Ischaemic heart disease.
Table 6 Relative risks and 95%CIs by current amount smoked1
StudyCigars per dayRR (95%CI) lung cancerRR (95% CI) COPDRR (95%CI) IHDAdjustment factors
CPS I[23]
1-20.90 (0.54 to 1.66)1.39 (0.74 to 2.38)0.98 (0.91 to 1.07)Age
3-42.36 (1.49 to 3.54)1.78 (0.89 to 3.18)1.06 (0.96 to 1.16)
5+3.40 (2.34 to 4.77)1.03 (0.37 to 2.23)1.14 (1.03 to 1.24)
CPS II[32]
Age 30-74 years11.18 (0.76 to 1.82)Age+2
2-31.43 (1.03 to 1.99)
4+1.33 (0.95 to 1.86)
Age 75+ years11.07 (0.64 to 1.78)
2-30.72 (0.45 to 1.16)
4+1.03 (0.70 to 1.51)
DORN[27]3
Age 55-64< 50.64 (0.15 to 2.69)0.84 (0.70 to 1.01)None
5+3.74 (1.53 to 9.10)0.98 (0.78 to 1.24)
Age 65-74< 51.32 (0.65 to 2.69)1.07 (0.94 to 1.23)
5+1.83 (0.78 to 4.27)1.21 (1.02 to 1.45)
KAISER[13]
< 51.57 (0.67 to 3.66)1.30 (0.93 to 1.81)1.20 (1.03 to 1.40)Age+4
5+3.24 (1.01 to 10.40)2.25 (1.39 to 3.65)1.56 (1.21 to 2.01)
KAUFMAN[33]
1-40.90 (0.30 to 2.70)Age+5
5+1.70 (0.60 to 4.80)
NLMS[17]
< 10.74 (0.08 to 7.26)Age+6
1+4.18 (2.34 to 7.46)
SADOW[30]7
11.62None
24.51
34.38
4+2.54
1Relative to never smoking.
2Adjusted also for education, hypertension, BMI, exercise, alcohol, environmental tobacco smoke and vitamin C.
3Estimated from data provided.
4For lung cancer and chronic obstructive pulmonary disease adjusted also for race, body mass index (BMI), history of diabetes, alcohol and occupational exposure; for ischaemic heart disease adjusted also for race, BMI, history of diabetes, alcohol, education, systolic blood pressure and serum cholesterol.
5Adjusted for religion, education, ethnicity, personality, family history of myocardial infarction, doctor visits, geography, physical activity, BMI, and treatment for diabetes, cholesterol, hypertension and chest pain.
6Adjusted also for sex, race, education and survey year.
795% CI not available, RRs estimated from data provided. RRs: Relative risks; COPD: Chronic obstructive pulmonary disease; IHD: Ischaemic heart disease.