Table 1.

Main characteristics of cohort studies included in the meta-analysis. [M=men; W=women; BMI=body mass index.]

Study	Country	Age (years)	Gender	Follow-up (years)	Partici-pants	Cases	Exposure definition & comparison	Case ascertainment	Adjustment for covariates	Study quality
Mutambudzi & Javed, 2016 (28)	US	≥50	M (43%) / W	7	1396	167	High strain, active job, and passive job vs low strain	Self-reported	BMI, physical activity, education, race, gender, alcohol use, average work hours/week, occupational category, marital status, insurance coverage, and hypertension	7
Huth et al, 2014 (29)	Germany	29-66	M (62.8%) / W	12.7	5337	291	High strain, active job, and passive job vs low strain	Self-reported and the date of diagnosis were validated by hospital records or physicians.	Sex, age, baseline survey, education, physical intensity of work, parental history of diabetes, living alone, physical inactivity, smoking, alcohol intake, BMI	9
Nyberg et al, 2014 (13)	European	44.1 (mean)	M (43.27%) / W	10.3	124 808	3703	High job demands and low control vs others	National health registers, clinical screening, and self-reports	Age, sex, and socioeconomic status	7
Kawakami et al, 1999 (22)	Japan	≥18	M	8	2194	34	High job demands and low control vs others	Medical check up	None	7
Heraclides et al, 2009 (26)	UK	35–55	M (70.67%) / W	11.6	5895	308	High job demands and low control vs others	Glucose tolerance test	Age	8
Kroenke et al, 2006 (27)	US	29–46	W	69.3 months	62 574	86	High strain, active job, and passive job vs low strain	Questionnaire and verified through medical records	Age ,BMI, family history of diabetes , work hours , rotating night-shift work , hours at work sitting, , job support, hours per week of work at home, leisure-time physical activity, smoking, alcohol intake, transunsaturated fat intake , glycemic load , caffeine intake , marital status, number of children, menopausal status , vitamin supplementation, and aspirin use	8
Norberg et al, 2007 (23)	Sweden	40-60	M (58.9%) / W	7.8	584	191	High strain, active job, and passive job vs low strain	Diagnosis registers	Age, sex and year of health survey	8
Eriksson et al, 2013 (24)	Sweden	35–56	M (41%) / W	8–10	5432	171	High strain, active job, and passive job vs low strain	Questionnaire and glucose tolerance test	Age, education, BMI, physical activity, smoking, family history of diabetes, and psychological distress	9
Pan et al, 2017 (25)	Sweden	≥60	M (35.4%) / W	6	2719	154	High job demands and low control vs others	Glycated haemoglobin level, self-report, hypoglycaemic medication use and clinical records	Sex, age, education level, vital status, follow-up time, BMI, smoking, alcohol consumption and physical activity, history of vascular diseases and the history of heart operation, cholesterol, blood pressure and depression	7