Systematic review with meta-analysis of the epidemiological evidence relating smoking to type 2 diabetes

Table 1 Literature searching

Step		Papers originally selected as probably relevant1	Papers rejected during data entry2	Papers providing separate results for multiple studies3
1	Previous reviews	98	10	3[9-11]
2	Medline search	74 (from 3365 hits)	23	4[12-15]
3	In-house database	1	0	0
4	Embase search	33 (from 5433 hits)	7	0
5	Secondary references4	15 (of 30 identified)	2	0
Total		221	42	7

¹Numbers of papers originally selected exclude those already identified in a previous step.

²Reasons for rejection are summarized in Supplementary file 1.

³Or for separate periods of follow-up.

⁴From papers identified in steps 2 to 4.

Table 2 Some characteristics¹ of the 145 studies of smoking and type 2 diabetes

Study Ref.	Main/ Other Ref.	Continent	Country, location2	Study Population3	Sex	Baseline	Follow-up (yr)4	Diagnosis code5	Cohort size	Diabetes cases	Age
3 studies6	[16]	North America	United States	Medical professionals	M+F	1984-1991	19.6	3	162807	12384	25-75
AICHI	[17]/[18]	Asia	Japan, Aichi	Civil servants	M+F	2002	9.0	3	3338	225	35-66
AIZAWA	[19]	Asia	Japan, Matsumoto	Participants from hospital (not otherwise defined)	M+F	2005	4.9	2	4159	279	Any
ALEIN	[20]	Asia	Taiwan (China), A-Lein	Persons undergoing community wide screening for hepatitis	M+F	1996-1997	8.0	2	3539	423	40-69
ALSWH	[21]	Oceania	Australia	General population	F	1998	12.0	1	12367	871	47-52
ANSAN	[22]/[23,24]	Asia	South Korea, Ansun and Ansan	Community based	M+F	2001-2002	4.0	2	4041	329	40-69
ARIC	[25]/[26]	North America	United States, North Carolina, Mississippi, Maryland	Probability sample from 4 US communities with exclusive sampling of African Americans in one of the four sites, Black or White	M+F	1987-1989	9.0	3	10892	1254	45-64
ASAN	[27]	Asia	South Korea, Asan	Attending voluntary medical check-ups	M+F	2000	5.0	2	5372	234	20-79
ATTICA	[28]	Europe	Greece, Athens	General population	M+F	2001-2002	10.0	2	1485	191	18-89
Ausdiab	[29]	Oceania	Australia	General population	M+F	1990-2000	5.0	2	5842	244	25+
BEDFORD	[30]	Europe	England, Bedford	Borderline diabetics with a 2h fasting glucose of 6.7-11.1 mmol/L	M+F	1962-1964	10.0	2	241	36	18+
BIP	[31]	Asia	Israel	Subjects with impaired functional capacity (New York Heart Association class II and III)	M+F	1990-1993	6.2	2	630	98	45-74
BMES	[32]	Oceania	Australia, West of Sydney	Non institutionalised residents	M+F	1992-1994	10.0	3	2123	165	49+
BOGALUSA	[33]	North America	United States, Bogalusa	General population	M+F	1973-2010	9.1	2	7725	176	<18
BOTNIA	[9]/[34]	Europe	Finland, Botnia	Family members of diabetics	M+F	1990	7.6	2	2770	138	Any
BRHS	[35]	Europe	Britain	General population	M	1978-1980	16.8	3	7124	290	40-59
BRUNECK	[36]	Europe	Italy, Bruneck	General population, White	M+F	1990	10.0	2	837	64	40-79
BURKE	[37]	Oceania	Australia Kimberley	General population, Aboriginal	M+F	1988-1989	12.9	2	493	104	15-88
BWHS	[38]	North America	United States	African American subscribers to magazine targeted at black women	F	1995	16.0	3	43003	4387	21-69
CASSAN	[39]	North America	United States	Majority were veterans, 98% Caucasian	M	1963	18.0	2	1972	226	20-80
CCHS	[40]	North America	United States, Cleveland	General population	M+F	2008	5.0	2	5084	872	18+
CDCdeC	[41]	Europe	Spain, Canaries	General population	M+F	2000-2005	3.5	3	5521	146	18-75
CEHSC	[42]	Asia	Hong Kong (China)	General population volunteers	M+F	1998-2001	9.8	2	53905	806	65+
CKB	[43]	Asia	China	General population	M+F	2004-2008	7.2	2	461211	8784	30-79
CoLaus	[44]	Europe	Switzerland, Lausanne	General population	M+F	2003-2006	5.5	2	3166	47	35-75
CPSI	[45]	North America	United States	General Population	M+F	1959-1960	12.0	3	709827	25397	30+
CRISPS	[46]	Asia	Hong Kong (China)	General population, Chinese	M+F	2000-2004	9.0	2	1380	123	Any
CURES	[47]	Asia	India, Chennai	General population	M+F	2001-2003	9.1	2	1376	385	20+
DAQING	[48]	Asia	China	Care clinic patients with pre-diabetes, part of diabetes prevention intervention	M+F	1986	23.0	3	568	436	Any
DEHGHA	[49]	Europe	Netherlands, Ommoord	General population	M+F	1990-1993	10.8	2	6935	645	55+
DE-PLAN	[11]	Europe	Spain, Navarra, Reus and Barcelona	Participants in clinical trial on Mediterranean diet, Caucasian	M+F	2006	4.2	2	552	124	45-75
DESIR	[50]	Europe	France, Western	Volunteers for periodic health checks	M+F	1998	9.0	2	3817	203	30-64
DLCS	[51]	Europe	Netherlands, Northern	General population, Western Europe	M+F	2007-2013	4.0	3	72880	1056	18-90
DNC	[52]	Europe	Denmark	Nurses	F	1993-1999	15.3	2	24174	1137	44+
DONGFENG	[53]	Asia	China, Da Qing	Retired employees	M+F	2008-2010	4.0	3	17690	1390	Any
DWECS	[54]	Europe	Denmark	Workers	M+F	1995-2005	5.0	2	6823	NA	30-59
EPIC-IN	[55]	Europe	8 countries7	Subset of participants in EPIC-InterAct cohort	M+F	1991	11.7	3	23501	10327	Any
ESTHER	[56]	Europe	Germany, Saarland	General population	M+F	2000-2002	8.0	3	7462	718	50-75
FAGERB	[57]	Europe	Sweden, Göteborg	General population, Caucasian	F	2001-2004	5.5	2	341	69	64
FINNMARK	[58]	Europe	Norway, Finnmark	General population	M+F	1997-1978	12.0	2	11654	162	35-52
GLOSTRUP	[59]	Europe	Denmark, Glostrup	General population	M	1982-2001	18.9	2	5350	211	30-70
GNHIES	[60]	Europe	Germany	General population (non institutionalized)	M+F	1997-1999	5.0	2	3625	82	18-79
HDNNCDS	[12]	Asia	China, Harbin	General population, Chinese	M+F	2010	4.2	3	7133	578	20-74
HEALTH2000	[10]	Europe	Finland	General population	M+F	2000-2001	7.0	2	4110	81	40-79
HEINZN	[61]/[62]	Europe	Germany, Western	General population	M+F	2000-2003	5.1	3	3547	319	45-75
HENAN	[63]	Asia	China, Henan	General population, N Chinese ancestry	M+F	2007-2008	6.0	3	12272	775	18+
HIPOP-OHP	[64]	Asia	Japan	Employees	M+F	1999	3.4	3	6498	229	Any
HIPPIS1	[65]	Europe	England and Wales	Primary care patients	M+F	1993-2008	8.0	2	2540753	78081	25-79
HIPPIS2	[66]	Europe	England	Primary care patients	M+F	2005-2016	3.9	2	8186705	178314	25-84
HISAYAMA	[67]	Asia	Japan, Hisayama	General population	M+F	1988	11.8	2	1935	286	40-79
HPFUS	[68]	North America	United States	Health professionals	M	1986	6.0	3	41810	509	40-75
HPHS	[12]	Asia	China, Harbin	General population, Chinese	M+F	2008	4.2	3	3350	244	20-74
HUNT	[69]	Europe	Norway, Nord-Trøndelag	General population	M+F	1984-1997	11.0	3	90819	1860	20+
ICARIA	[70]	Europe	Spain	Spanish workers	M+F	2004-2007	4.1	3	380366	9960	18-65
ICS	[71]	Asia	Iran, Isfahan, Arak and Najafabad	General population	M+F	2001	7.0	2	2980	389	35+
IPC	[72]	Europe	France, Paris	Workers and those seeking employment who had undergone 2 health checks	M+F	1998-2010	5.3	2	22567	527	18+
IRAS	[73]	North America	United States, 4 areas8	General population	M+F	1992-1993	5.0	2	906	148	40-69
IWHS	[74]	North America	United States, Iowa	Community based	F	1986	13.2	1	36839	3281	55-69
JACC	[75]	Asia	Japan	Community based	M+F	1988-1990	5.0	1	16160	396	40-79
J-ECOH	[76]/[77]	Asia	Japan	Employees	M+F	2008-2010	3.9	2	53930	2441	15-83
JHS	[78]	North America	United States, Mississippi	General population, Black	M+F	2000-2004	8.0	2	2991	479	21-84
JPHC	[79]	Asia	Japan	General population	M+F	1990	10.0	1	28893	1183	40-59
JPHC2	[80]	Asia	Japan	General population	M+F	1995-1998	5.0	1	59834	1100	45-74
KANGBUK	[81]	Asia	South Korea, Seoul	Individuals undergoing health screening	M+F	2002	5.6	3	174314	5544	18+
KAWAHA	[82]	Asia	Japan, Kitakyushu City	City workers	M+F	2008	3.7	2	52781	4369	20-89
KAWAKA	[83]	Asia	Japan, electrical company	Employees of large electrical company	M	1984	8.0	2	2312	41	18-53
KMIC	[84]	Asia	South Korea	Government and school employees	M	1990-1986	8.0	2	27635	1170	35-44
KoGES-K	[85]/[86]	Asia	South Korea, Kangwha	Community based	M+F	2006-2011	4.0	2	2079	142	40+
KORA F4/FF4	[87]	Europe	Germany, Augsburg	General population	M+F	2006-2008	7.0	2	504	76	62-81
KORA S4/F4	[88]	Europe	Germany, Augsburg	General population	M+F	1999-2001	7.0	2	887	93	55-74
KPNW	[89]	North America	United States, Portland	Health care members	M+F	1997-2000	6.8	2	46578	1854	40+
LEICESTER	[90]	Europe	England, Leicester	With clinical diagnosis of polycystic ovary syndrome	F	1988-2009	5.2	2	2164	138	16-79
LIETO	[91]	Europe	Finland, Leito	General population	M	1998-1999	9.0	2	430	30	64+
LINDBE	[92]	Europe	Denmark, Copenhagen	General population	M+F	2001-2003	8.5	2	5349	136	20-94
LLP	[93]	Europe	England, Liverpool	General population	M+F	1998-2008	10.0	2	8753	763	45-79
MAILES	[94]	Oceania	Australia, Adelaide	General population	M	2002-2006	4.9	3	1597	232	35-80
MANSON	[95]	North America	United States	Physicians in randomized trial	M	1982	12.0	1	21068	770	40-84
MECC	[96]	North America	United States, Hawaii and California	General population, African American and Latino	M+F	1993-1996	14.0	3	48995	15833	50-75
MECH	[97]	North America	United States, Hawaii	General population, Caucasian, Hawaiian, Japanese, American	M+F	1993-1996	12.1	3	74970	8559	45-75
MESA	[98]/[99]	North America	United States, 6 states9	General population, White, Black, Hispanic or Chinese	M+F	2000-2002	10.2	2	5931	359	45-84
MFH	[10]	Europe	Finland	General population	M+F	1978-1980	10.0	2	4517	145	40-79
MJH	[100]	Asia	Taiwan (China)	Paid members of private health screening program, Chinese	M+F	2001-2014	6.7	3	147908	4781	18+
MONICAG	[101]	Europe	Germany, Augsburg	General population	M+F	1984-1995	12.5	3	10892	672	25-74
MONICAS	[102]	Europe	Sweden, Northern	General population	M	1986-1994	8.7	3	1275	27	25-74
MORIMO	[103]/[104]	Asia	Japan, Nagano prefecture	Volunteers in Nagano Prefecture	M+F	1990-1992	10.1	3	5872	595	40-69
MOZAFF	[105]	North America	United States, 4 states10	Ambulatory, noninstitutionalized subjects	M+F	1989-1992	10.0	2	4883	337	65+
MPBB	[106]	North America	United States, Michigan	Subjects who had injected food contaminated with polybrominated biphenyls, 99.8% White	M+F	1976	25.0	3	1384	180	20+
MPP	[9]	Europe	Sweden, Malmo	General population	M+F	1974-1992	24.8	2	16061	2063	Any
MUTUAL	[107]	Asia	Japan	Civil servants	M+F	2000	6.5	2	5848	287	30-59
MYHUS	[108]	Asia	Japan	Employees	M+F	2004	5.0	3	13700	408	36-55
NAGALA	[109]/[110]	Asia	Japan, Gifu	Subjects receiving medical check-ups	M+F	2004-2015	5.1	3	17810	804	Any
NAGAYA	[111]	Asia	Japan, Nagoya	Volunteer attendants of annual health check ups	M	1988-1990	7.4	3	16829	869	30-59
NAKANI	[112]	Asia	Japan, Osaka	Employees	M	1994	5.0	2	1266	54	35-59
NCDS	[113]	Europe	Britain	Birth cohort from March 1958	M+F	1974	17.0	1	4945	28	16
NHANES	[114]	North America	United States	General population	M+F	1971-1975	18.0	3	4830	443	25-74
NHIC	[115]	Asia	South Korea	Recipients of biennial medical check-ups	M+F	1992-1995	14.0	2	1236443	89422	30-95
NHIS-HEALS	[116]	Asia	South Korea	Recipients of national health screen test	M+F	2002-2003	10.8	2	359349	37678	40-79
NHIS-NCS	[117]	Asia	South Korea	Nationally representative	M+F	2002	6.8	2	51405	2749	20+
NHS	[118]	North America	United States	Registered Nurses	F	1976-1982	24.0	3	100526	5392	30-55
NHSII	[13]	North America	United States	Registered Nurses	F	1989-1991	23.0	3	88086	5441	25-42
NIH -AARP	[119]	North America	United States, 6 states11	General population	M+F	1995-1996	11.0	1	207479	18000	50-71
NOMAS	[120]	North America	United States, North Manhattan	General population, White, Black or Hispanic	M+F	1993-2001	11.0	3	2430	449	40+
NOVAK	[121]	Europe	Sweden, Gothenburg	General population (intervention group in ineffective trial)	M	1970-1973	35.0	2	6828	899	47-56
OLMSTED	[122]	North America	United States, Rochester	General population who also took at least one medication	M+F	1999-2004	6.0	2	13508	1182	18+
ONAT	[123]	Asia	Turkey	Participants in nationwide survey	M+F	1997-1998	5.9	3	3385	216	28+
OSAKA	[124]	Asia	Japan, Osaka	General population undergoing basic health check-ups	M+F	2001	4.0	2	9327	171	40-74
OSLO	[125]	Europe	Norway, Oslo	General population	M	1972-1973	28.0	3	6382	584	40-49
OSTENS	[126]	Europe	Sweden, Stockholm	General population	M	1992-1994	10.0	2	2383	99	35-56
PARK	[127]	Asia	South Korea, not known	Undergoing health examinations	M	2002	4.0	2	1717	50	Any
PATJA	[128]	Europe	Finland, North Karelia and Kuopio	General population	M+F	1972-1992	21.0	2	41372	2770	25-64
PINGLIANG	[129]	Asia	China, Ping Liang	General population pre-diabetic at baseline	M+F	2002-2003	10.8	2	334	98	Any
PMMJS	[130]	Asia	China, Jiangsu	General population	M+F	2000-2004	5.0	2	3598	160	35-74
PREDIMED	[11]	Europe	Spain, Navarra, Reus and Barcelona	Participants in clinical trial on Mediterranean diet, Caucasian	M+F	2003-2009	4.8	2	1381	155	55-80
PREDIMERC	[131]	Europe	Spain, Madrid	General population	M+F	2007	6.4	2	2048	44	30-74
PREVEND	[132]	Europe	Netherlands, Groningen	General population	M+F	1997-1998	11.4	3	7953	447	Any
REGARDS	[133]	North America	United States	General population, Black or White	M+F	2003-2007	9.5	2	7758	891	45+
SABE	[134]	South America	Brazil, Sào Paulo	General population	M+F	2000	6.0	1	914	72	60+
SAIREN	[135]	Asia	Japan, Ibaraki-ken	General population undergoing annual health check-ups	M+F	1993	5.0	2	128141	7990	40-79
SALSA	[136]	North America	United States, Sacramento	General population, Latino	M+F	1998-1999	10.0	3	782	144	60+
SAMSUNG	[137]	Asia	South Korea, Seoul	Undergoing health examinations, Korean	M	2006	6.0	3	1774	180	20+
SAPALDIA	[138]	Europe	Switzerland	General population	M+F	2002	8.3	3	2631	110	18+
SAWADA	[139]	Asia	Japan, Tokyo	Employees of Tokyo Gas Company	M	1985	14.0	3	4745	280	20-41
SAX45	[140]	Oceania	Australia, New South Wales	General population	M+F	2006-2008	3.4	1	54997	888	45+
SCCS	[14]	North America	United States, Southern	General population, Black or White	M+F	2002-2009	4.5	1	35892	3439	40-79
SCCS2	[14]	North America	United States, Southern	General population, Black or White	M+F	201212	3.0	1	20712	1708	43-82
SHFS	[141]	North America	United States, 4 states13	Members of multiplex families, American Indians	M+F	2001-2003	5.5	2	431	133	14+
SHIP	[142]	Europe	Germany, Augsburg	Caucasian German citizens	M+F	1997-2001	11.1	2	2034	206	20-81
SMHS	[143]	Asia	China, Shanghai	General population	M	2002-2006	5.4	3	51464	1304	40-74
STILLW	[144]	Europe	Finland	Employees of Finnish Company	M	1986	17.0	2	5827	313	18-65
STRAND	[145]	Europe	Finland, Helsinki	Volunteer executives and businessmen	M	1974-1975	20.0	3	1802	94	40-56
STRING	[146]/[147]	Europe	England, London	Civil service employees	M+F	1985-2002	23.7	2	8270	1286	50
SUGIMO	[148]	Asia	Japan, Tokyo	Participants in MHTS	M+F	1976	16.0	2	2573	296	18-69
SULAWESI	[149]	Asia	Indonesia, South Sulawesi	Three tribes	M+F	2013	3.0	2	182	58	16+
SWAN	[150]	North America	United States, Michigan	Participants in study of menopause transition, Black or White	F	1996	16.0	3	424	157	42-52
TCS	[151]	Asia	Thailand	Students at Sukothai Thammithirat Open University	M+F	2005	8.0	1	39507	698	15-88
TERATA	[152]	Asia	Japan, Chiba	Steelworkers	M	2002	8.0	2	8423	464	Any
TLSA	[153]	Asia	Taiwan (China), Non-aboriginal areas	Participants in ongoing survey on aging, Taiwanese	M+F	1999	4.0	1	2995	277	53+
TOPICS6	[154]	Asia	Japan, Toranomon	Government employees and some general population	M+F	1997-2002	5.0	3	7654	289	40-75
TROMSO	[155]	Europe	Norway, Tromsø	General population	M+F	1994-1995	10.8	3	26168	522	25-98
UCHIMO	[156]	Asia	Japan, Osaka	Employees of large company	M	1981-1991	10.0	2	6250	450	35-60
VETERAN	[157]	North America	United States	Veterans	M+F	2002-2003	4.0	2	239057	33453	18-99
VIP	[158]	Europe	Sweden, Västerbotten County	General population	M+F	1990-2012	9.9	3	32120	2211	35-55
WHI	[159]	North America	United States	Postmenopausal women in a clinical trial or an observational study	F	1993-1998	11.0	1	135906	15076	50-79
YOUNGF	[160]	Europe	Finland	Population based	M+F	1980	24.0	3	2298	79	3-18
ZUTPHEN	[161]	Europe	Netherlands, Zutphen	General population	M	1960	25.0	2	841	58	40-59

¹Where relevant, characteristics are shown for the main reference, given first in the column Main/Other Ref.

²If location not stated, then national.

³All races are included unless stated otherwise.

⁴NA means not available. Some studies provided results for more than one follow-up time. Here the longest follow-up is indicated. The follow-up times are presented as means, medians or averages to various numbers of decimal places. The values shown are the best estimate available.

⁵1 = self-report only; 2 = medical records only; 3 = both.

⁶Studies HPFUS, NHS and NHSII.

⁷France, Italy, Spain, United Kingdom, Netherlands, Germany, Sweden and Denmark.

⁸Los Angeles, Oakland, San Antonio and San Juis Valley.

⁹Maryland, Illinois, North Carolina, California, New York and Minnesota.

¹⁰North Carolina, California, Maryland and Pennsylvania.

¹¹California, Florida, Louisiana, New Jersey, North Carolina and Pennsylvania.

¹²Subset of SCCS who were diabetes free at end of SCCS follow-up. Unclear what the baseline date range of SCCS2 actually was.

¹³Arizona, North and South Dakota and Oklahoma. M: Male; F: Female.

Table 3 Meta-analysis random effect relative risks for current smoking

Grouping1		Current vs never smoking			Current vs non-smoking
		n2	RR (95%CI)	P	n	RR (95%CI)	P
Overall		99	1.33 (1.28-1.38)	P < 0.001, P < 0.05	156	1.28 (1.24-1.32)	P < 0.001, P < 0.05
Sex	Female	25	1.30 (1.23-1.37)		31	1.26 (1.21-1.31)
	Male	36	1.40 (1.32-1.49)		47	1.30 (1.24-1.36)
	Combined	38	1.28 (1.18-1.39)	NS3	78	1.26 (1.18-1.34)	NS
Continent	Asia	44	1.36 (1.30-1.43)		57	1.36 (1.29-1.43)
	Europe	32	1.34 (1.27-1.42)		60	1.25 (1.20-1.30)
	North and South America	19	1.27 (1.18-1.37)		34	1.18 (1.12-1.25)
	Oceania	4	1.05 (0.68-1.62)	NS	5	1.54 (1.28-1.85)	P < 0.001
Publication year	Up to 2005	13	1.41 (1.27-1.56)		23	1.24 (1.16-1.33)
	2005-2014	47	1.36 (1.30-1.43)		66	1.31 (1.27-1.35)
	2015 or later	39	1.27 (1.20-1.35)	NS	67	1.23 (1.17-1.30)	NS
Basis of diagnosis	Self-report only	12	1.32 (1.25-1.40)		17	1.34 (1.25-1.44)
	Medical records only	49	1.32 (1.25-1.38)		86	1.29 (1.23-1.34)
	Both	38	1.36 (1.27-1.46)	NS	53	1.24 (1.17-1.32)	NS
Population	General	93	1.32 (1.28-1.37)		147	1.28 (1.24-1.32)
	Pre-diabetics only	2	3.29 (1.51-7.21)		3	1.23 (0.79-1.90)
	Pre-diabetics excluded	4	1.61 (1.30-1.99)	P < 0.05	6	1.38 (1.15-1.67)	NS
Number of adjustment factors	0	17	1.15 (1.00-1.33)		33	1.19 (1.08-1.31)
	1 to 5	18	1.36 (1.25-1.47)		30	1.38 (1.27-1.51)
	6 to 10	43	1.40 (1.32-1.48)		64	1.29 (1.25-1.33)
	11 or more	21	1.28 (1.20-1.37)	P < 0.05	29	1.22 (1.15-1.30)	P < 0.1
Cohort size	< 5000	35	1.36 (1.19-1.56)		58	1.31 (1.20-1.42)
	5000 to 20000	20	1.38 (1.25-1.53)		43	1.24 (1.17-1.32)
	> 20000	44	1.32 (1.26-1.37)	NS	55	1.29 (1.24-1.35)	NS
Number of type 2 diabetes cases	< 500	44	1.37 (1.23-1.52)		78	1.27 (1.19-1.35)
	500-999	18	1.50 (1.34-1.67)		24	1.40 (1.27-1.55)
	1000-2000	10	1.26 (1.15-1.38)		17	1.20 (1.11-1.30)
	2001+	27	1.29 (1.22-1.35)	P < 0.1	37	1.26 (1.20-1.33)	NS
Highest age at baseline	< 60	13	1.36 (1.23-1.51)		22	1.24 (1.16-1.32)
	60-74	27	1.44 (1.32-1.56)		38	1.36 (1.27-1.45)
	75+	59	1.29 (1.24-1.35)	P < 0.1	96	1.26 (1.21-1.31)	NS
Length of follow-up (yr)	< 5	14	1.27 (1.19-1.35)		25	1.24 (1.15-1.34)
	5-10	55	1.38 (1.30-1.47)		81	1.34 (1.28-1.40)
	> 10	30	1.31 (1.22-1.39)	NS	50	1.22 (1.17-1.28)	P < 0.05
Definition of smoking	Cigarette	47	1.32 (1.27-1.38)		63	1.25 (1.21-1.29)
	Smoking	50	1.36 (1.26-1.46)		89	1.30 (1.23-1.37)
	Tobacco	2	1.10 (0.94-1.29)	P < 0.1	4	1.16 (1.06-1.27)	P < 0.1
Adjusted for age	No	20	1.17 (1.04-1.32)		41	1.22 (1.12-1.33)
	Yes	79	1.35 (1.31-1.41)	P < 0.05	115	1.29 (1.25-1.33)	NS
Adjusted for sex	No	72	1.35 (1.29-1.41)		107	1.27 (1.23-1.32)
	Yes	27	1.29 (1.20-1.39)	NS	49	1.29 (1.20-1.38)	NS
Adjusted for BMI	No	29	1.24 (1.11-1.38)		55	1.22 (1.13-1.32)
	Yes	70	1.35 (1.30-1.41)	NS	101	1.30 (1.26-1.34)	NS
Adjusted for physical activity	No	41	1.27 (1.20-1.35)		87	1.27 (1.21-1.33)
	Yes	58	1.36 (1.30-1.43)	P < 0.1	69	1.28 (1.23-1.33)	NS
Adjusted for alcohol consumption	No	42	1.26 (1.19-1.34)		87	1.26 (1.20-1.32)
	Yes	57	1.37 (1.31-1.43)	P < 0.05	69	1.29 (1.25-1.33)	NS
Adjusted for family history of diabetes	No	61	1.28 (1.22-1.35)		99	1.23 (1.17-1.29)
	Yes	38	1.41 (1.33-1.49)	P < 0.05	57	1.34 (1.29-1.40)	P < 0.01
Adjusted for education	No	63	1.37 (1.31-1.44)		115	1.29 (1.24-1.35)
	Yes	36	1.28 (1.21-1.34)	P < 0.05	41	1.23 (1.18-1.28)	P < 0.1
Adjusted for diet	No	74	1.35 (1.29-1.41)		126	1.29 (1.24-1.34)
	Yes	25	1.30 (1.22-1.38)	NS	30	1.23 (1.18-1.28)	P < 0.1
Adjusted for blood pressure	No	53	1.31 (1.24-1.40)		88	1.27 (1.21-1.34)
	Yes	46	1.35 (1.29-1.41)	NS	68	1.28 (1.24-1.33)	NS
Adjusted for cholesterol	No	72	1.30 (1.25-1.35)		115	1.26 (1.22-1.31)
	Yes	27	1.40 (1.32-1.48)	P < 0.05	41	1.32 (1.25-1.39)	NS
Adjusted for glucose	No	79	1.30 (1.25-1.35)		116	1.26 (1.22-1.31)
	Yes	20	1.44 (1.35-1.54)	P < 0.01	40	1.34 (1.27-1.41)	NS
Adjusted for triglycerides	No	80	1.30 (1.25-1.36)		124	1.27 (1.22-1.31)
	Yes	19	1.45 (1.33-1.58)	P < 0.05	32	1.34 (1.24-1.44)	NS
Adjusted for waist circumference	No	82	1.34 (1.29-1.40)		136	1.28 (1.24-1.32)
	Yes	17	1.29 (1.19-1.41)	NS	20	1.25 (1.16-1.35)	NS
Adjusted for any other factors	No	37	1.30 (1.19-1.42)		62	1.28 (1.18-1.38)
	Yes	62	1.34 (1.29-1.40)	NS	94	1.27 (1.23-1.30)	NS

¹For sex, publication year, basis of diagnosis, number of adjustment factors, definition of smoking and age adjusted the grouping relates to characteristics of the relative risk. For other factors it relates to characteristics of the study.

²Number of estimates combined.

³NS means not significant, P ≥ 0.1. For the overall analysis, the first P value relates to heterogeneity between estimates and the second to publication bias. For the other analyses it relates to a test of heterogeneity between the random-effects estimates at each level. Information on publication bias by level of each factor studied is given in Supplementary file 6. CI: Confidence interval; RR: Relative risk; NS: Not significant; BMI: Body mass index.

Table 4 Dose-response analyses for current smoking

Grouping1	Current vs never smoking		Current vs non-smoking
	n2	RR (95%CI)	n	RR (95%CI)
Using key values:
About 10 cigs/d	13	1.10 (1.03-1.18)	13	1.04 (0.98-1.10)
About 20 cigs per d	13	1.31 (1.19-1.44)	13	1.27 (1.16-1.39)
About 40 cigs per d	13	1.55 (1.39-1.72)	13	1.54 (1.37-1.72)
Low	23	1.17 (1.11-1.23)	22	1.13 (1.07-1.19)
Medium	23	1.30 (1.22-1.39)	22	1.26 (1.18-1.34)
High	23	1.53 (1.41-1.65)	22	1.48 (1.37-1.60)
1-19 cigs/d	18	1.32 (1.20-1.45)	17	1.20 (1.10-1.30)
20+	18	1.58 (1.42-1.76)	17	1.44 (1.31-1.59)

¹The key value analysis is based on all studies which provide estimates for each key value (i.e., for a range which included the key value and no other key value). The low/medium/high analysis is based on all studies which provide estimates for exactly three levels. The 1-19, 20+ analysis is based on those studies which reported results only for these two levels.

²Number of estimates combined. CI: Confidence interval; RR: Relative risk.

Table 5 Meta-analysis random effects relative risks for former (vs never) smoking

Grouping1		n2	RR (95%CI)	P
Overall		100	1.13 (1.11-1.16)	P < 0.001, P < 0.01
Sex	Female	26	1.13 (1.08-1.18)
	Male	36	1.12 (1.08-1.16)
	Combined	38	1.16 (1.09-1.22)	NS3
Continent	Asia	44	1.16 (1.10-1.22)
	Europe	32	1.13 (1.09-1.18)
	North and South America	20	1.11 (1.07-1.16)
	Oceania	4	1.07 (0.93-1.23)	NS
Publication year	Up to 2005	13	1.13 (1.06-1.21)
	2005-2014	47	1.16 (1.11-1.22)
	2015 or later	40	1.11 (1.08-1.15)	NS
Basis of diagnosis	Self-report only	12	1.17 (1.05-1.29)
	Medical records only	49	1.11 (1.08-1.13)
	Both	39	1.16 (1.11-1.22)	P < 0.1
Population	General	94	1.13 (1.11-1.16)
	Pre-diabetics only	2	0.97 (0.08-12.64)
	Pre-diabetics excluded	4	1.11 (0.86-1.44)	NS
Number of adjustment factors	0	18	1.11 (1.01-1.23)
	1 to 5	18	1.20 (1.11-1.30)
	6 to 10	42	1.12 (1.08-1.17)
	11 or more	22	1.13 (1.09-1.17)	NS
Cohort size	< 5000	35	1.21 (1.11-1.32)
	5000 to 20000	20	1.19 (1.09-1.29)
	> 20000	45	1.12 (1.09-1.15)	NS
Number of type 2 diabetes cases	< 500	44	1.21 (1.12-1.30)
	500 to 999	18	1.11 (1.03-1.20)
	1000 to 2000	10	1.26 (1.10-1.45)
	2001+	28	1.11 (1.08-1.14)	P < 0.1
Highest age at baseline	< 60	14	1.20 (1.10-1.30)
	60-74	27	1.19 (1.10-1.29)
	75+	59	1.11 (1.09-1.14)	NS
Length of follow-up (yr)	< 5	14	1.13 (1.08-1.19)
	5-10	55	1.16 (1.10-1.23)
	> 10	31	1.11 (1.08-1.15)	NS
Definition of smoking	Cigarette	48	1.12 (1.09-1.15)
	Smoking	50	1.15 (1.10-1.21)
	Tobacco	2	0.95 (0.83-1.08)	P < 0.05
Adjusted for age	No	21	1.13 (1.05-1.22)
	Yes	79	1.13 (1.10-1.16)	NS
Adjusted for sex	No	75	1.13 (1.10-1.16)
	Yes	25	1.13 (1.07-1.19)	NS
Adjusted for BMI	No	31	1.15 (1.07-1.24)
	Yes	69	1.12 (1.10-1.15)	NS
Adjusted for physical activity	No	41	1.15 (1.11-1.20)
	Yes	59	1.12 (1.09-1.16)	NS
Adjusted for alcohol consumption	No	43	1.15 (1.10-1.19)
	Yes	57	1.13 (1.09-1.16)	NS
Adjusted for family history of diabetes	No	61	1.13 (1.10-1.17)
	Yes	39	1.13 (1.09-1.17)	NS
Adjusted for education	No	65	1.16 (1.12-1.19)
	Yes	35	1.09 (1.05-1.14)	P < 0.05
Adjusted for diet	No	75	1.14 (1.11-1.17)
	Yes	25	1.12 (1.07-1.16)	NS
Adjusted for blood pressure	No	54	1.14 (1.10-1.19)
	Yes	46	1.13 (1.09-1.16)	NS
Adjusted for cholesterol	No	73	1.13 (1.10-1.16)
	Yes	27	1.14 (1.08-1.20)	NS
Adjusted for glucose	No	80	1.13 (1.10-1.16)
	Yes	20	1.15 (1.07-1.23)	NS
Adjusted for triglycerides	No	81	1.12 (1.10-1.15)
	Yes	19	1.17 (1.08-1.27)	NS
Adjusted for waist circumference	No	83	1.13 (1.10-1.16)
	Yes	17	1.14 (1.05-1.24)	NS
Adjusted for other factors	No	38	1.15 (1.08-1.23)
	Yes	62	1.13 (1.10-1.15)	NS

¹For sex, publication year, basis of diagnosis, number of adjustment factors, definition of smoking and age adjusted the grouping relates to characteristics of the RR. For other factors it relates to characteristics of the study.

²Number of estimates combined.

³NS means not significant, P ≥ 0.1. For the overall analysis, the first P value relates to heterogeneity between estimates and the second to publication bias. For the other analyses it relates to a test of heterogeneity between the random-effects estimates at each level. Information on publication bias by level of each factor studied is given in Supplementary file 6. CI: Confidence interval; RR: Relative risk; NS: Not significant; BMI: Body mass index.

Table 6 Dose-response analyses for former vs never smoking (years quit)

Grouping1		n2	RR (95%CI)
Using key values:	About 3 yr quit	8	1.39 (1.21-1.60)
	About 7 yr quit	8	1.17 (1.07-1.27)
	About 12 yr quit	8	1.11 (1.01-1.22)
Shortest		14	1.46 (1.31-1.63)
Longest		14	1.13 (1.01-1.27)

¹The key value analysis is based on all studies which provide estimates for each key value (i.e., for a range which included the key value and no other key value). The shortest/longest analysis is based on all studies which provide estimates by years quit.

²Number of estimates combined. CI: Confidence interval; RR: Relative risk.

Table 7 Meta-analysis random effects relative risks for ever (vs never) smoking

Grouping1		n2	RR (95%CI)	P
Overall		100	1.25 (1.21-1.28)	P < 0.001, P < 0.01
Sex	Female	24	1.25 (1.18-1.31)
	Male	36	1.25 (0.20-1.31)
	Combined	40	1.22 (1.14-1.31)	P < 0.05
Continent	Asia	41	1.30 (1.25-1.36)
	Europe	36	1.21 (1.17-1.26)
	North and South America	20	1.19 (1.13-1.26)
	Oceania	3	0.87 (0.48-1.57)	P < 0.05
Publication year	Up to 2005	13	1.25 (1.16-1.34)
	2005-2014	47	1.26 (1.20-1.33)
	2015 or later	40	1.23 (1.18-1.28)	NSc
Basis of diagnosis	Self-report only	10	1.35 (1.17-1.56)
	Medical records only	51	1.22 (1.18-1.27)
	Both	39	1.26 (1.19-1.33)	NS
Population	General	95	1.24 (1.21-1.28)
	Pre-diabetics only	1	3.30 (1.24-8.77)
	Pre-diabetics excluded	4	1.43 (1.17-1.76)	P < 0.1
Number of adjustment factors	0	23	1.18 (1.06-1.32)
	1 to 5	16	1.28 (1.20-1.36)
	6 to 10	40	1.24 (1.19-1.30)
	11 or more	21	1.22 (1.16-1.28)	NS
Cohort size	< 5000	39	1.26 (1.14-1.38)
	5000 to 20000	17	1.27 (1.17-1.38)
	> 20000	44	1.24 (1.20-1.28)	NS
Number of type 2 diabetes cases	< 500	46	1.26 (1.16-1.36)
	500 to 999	17	1.32 (1.19-1.47)
	1000 to 2000	9	1.28 (1.14-1.43)
	2001+	28	1.22 (1.17-1.26)	NS
Highest age at baseline	< 60	13	1.35 (1.23-1.47)
	60-74	27	1.32 (1.23-1.41)
	75+	60	1.21 (1.17-1.25)	P < 0.05
Length of follow-up (yr)	< 5	14	1.21 (1.15-1.26)
	5-10	56	1.29 (1.22-1.35)
	> 10	30	1.21 (1.15-1.28)	NS
Definition of smoking	Cigarette	48	1.22 (1.18-1.26)
	Smoking	50	1.28 (1.20-1.36)
	Tobacco	2	1.09 (0.94-1.25)	P < 0.1
Adjusted for age	No	27	1.19 (1.09-1.31)
	Yes	73	1.24 (1.20-1.28)	NS
Adjusted for sex	No	77	1.26 (1.22-1.30)
	Yes	23	1.20 (1.13-1.27)	NS
Adjusted for BMI	No	35	1.23 (1.13-1.34)
	Yes	65	1.24 (1.20-1.28)	NS
Adjusted for physical activity	No	43	1.26 (1.20-1.32)
	Yes	57	1.24 (1.19-1.29)	NS
Adjusted for alcohol consumption	No	46	1.24 (1.19-1.30)
	Yes	54	1.25 (1.20-1.30)	NS
Adjusted for family history of diabetes-	No	62	1.22 (1.17-1.27)
	Yes	38	1.28 (1.23-1.33)	NS
Adjusted for education	No	67	1.29 (1.24-1.34)
	Yes	33	1.17 (1.12-1.21)	P < 0.001
Adjusted for diet	No	76	1.26 (1.22-1.31)
	Yes	24	1.21 (1.15-1.26)	NS
Adjusted for blood pressure	No	57	1.25 (1.19-1.32)
	Yes	43	1.23 (1.19-1.27)	NS
Adjusted for cholesterol	No	75	1.24 (1.20-1.28)
	Yes	25	1.27 (1.19-1.36)	NS
Adjusted for glucose	No	79	1.23 (1.19-1.27)
	Yes	21	1.31 (1.25-1.37)	P < 0.05
Adjusted for triglycerides	No	83	1.23 (1.20-1.27)
	Yes	17	1.31 (1.22-1.41)	NS
Adjusted for waist circumference	No	84	1.25 (1.21-1.30)
	Yes	16	1.21 (1.12-1.31)	NS
Adjusted for other factors	No	42	1.24 (1.15-1.33)
	Yes	58	1.23 (1.20-1.28)	NS

¹For sex, publication year, basis of diagnosis, number of adjustment factors, definition of smoking and age adjusted the grouping relates to characteristics of the relative risk. For other factors it relates to characteristics of the study.

²Number of estimates combined.

³NS means not significant, P ≥ 0.1. For the overall analysis, the first P value relates to heterogeneity between estimates and the second to publication bias. For the other analyses it relates to a test of heterogeneity between the random-effects estimates at each level. Information on publication bias by level of each factor studied is given in Supplementary file 8. CI: Confidence interval; RR: Relative risk; NS: Not significant.