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

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.