Table 1.
Characteristics and results of the 21 studies. [BJSQ=Brief Job Stress Questionnaire; ERI=effort–reward imbalance; FPG=fasting plasma glucose; FU=follow up; JCQ=job control questionnaire; NI=not informed; OGTT=oral glucose tolerance test; PB=participation at baseline; PCE=prevalent cases excluded; PI=proportion included.]
| Study and country | Population characteristics Years of FU Type of workers N analyzed / eligible N women / men | Work-related stressors Measurement time Tool Exposed fraction | Diabetes cases (%) | Analyses / Results Model PCE: Yes/No/Unclear Covariates Results |
|---|---|---|---|---|
| Eriksson et al, 2013 (43) Sweden (SDPP) |
Baseline: 1992–1994 (men) or 1996–1998 (women) FU: 8-10 y Middle-aged workers: 4580/7949 PB: 72%, PI:80% 3205 / 2227 Mean age: 47.4 y |
Exposure at baseline, JCQ Demands: 5 items Job control: 5 items Social support: 2 items 22% high strain quadrant |
OGTT level: 171 cases (3%) 149 cases in N analyzed (3.2%) |
Logistic regression, OR (95% CI) PCE: Yes, but N unknown Age, sex, educational level, psychological distress Demands: 0.7 (0.5–1.1) Low job control: 1.2 (0.8–1.9) Job strain tertiles: 1.6 (1.0–2.6) High vs. Low strain: 0.8 (0.5–1.3) |
| Garbarino et al, 2018 (34) Italy |
Baseline: 2009, FU: 5y, Police (men) 234 / 294 PB: 99%, PI: 80% Mean age: 41 y |
Repeated exposure JCQ: Demands: 5 items; Job control: 6 items; Social support: 6 items 89% job strain ERI: Effort: 6 items; Reward: 11 items 13% high ERI |
FPG>100 mg/dL (5.6 mmol/L). 3 cases (1.3%) | Logistic regression, OR (95% CI) PCE: Yes Age, education, origin, marital status, housing and presence of offspring High combine stress: 6.36 (0.56–72.45) |
| Gilbert-Ouimet et al, 2021 (13) Canada (CCHS) |
Baseline: 2009 FU: 13.5y, Workers: 12 896 / NI PB: 63%, PI: 95% 6148/6749 Mean age: 41 y |
Exposure at baseline JCQ: Demands: 2 items; Job control: 5 items ♂ 36.3%, ♀ 26.1% all Highest job strain tertiles |
Administrative data, hospital records or two physician service claims | Cox regression model, HR (95% CI) PCE: Yes Stratified by or controlled for sex, age, marital status, children under 12 in the house, born in Canada, ethnicity, living location, immigration status, survey year, self– reported, chronic diseases, work hours, interview method and activity restrictions at work. ♂ high strain: 0.93 (0.69–1.27) ♀ high strain: 1.23 (0.86–1.77) |
| Heraclides et al, 2009 (38) United Kingdom (Whitehall II) |
Baseline: 1991–1993 FU 15y, Workers: 5895/10 308 PB: 73% (Heraclides 2012), PI 82% 1729/ 4166 Mean age: 48 y |
Exposure at baseline JCQ: Demands: 4 items; Job control: 15 items; Social support: 6 items High demands: ♂53%, ♀47%, all 51% Low control: ♂44%, ♀67%, all 51% Low support: ♂33%, ♀34%, all 33% Job strain: ♂ 24%, ♀30%, all 25% Isostrain: ♂ 11%, ♀14%, all 12% |
OGTT level and self-report 308 cases (5.2%) | Cox regression model, HR (95% CI) PCE: Yes Stratified by sex, controlled by age Demands: ♂106/2.222, 0.82 (0.63–1.07) ♀: 40/756, 1.06 (0.70–1.62) All: 146/2.978, 0.88 (0.70–1.10) Job control: ♂: 84/1846, 0.86 (0.66–1.13) ♀: 63/1171, 1.09 (0.70–1.69) All: 147/3017, 0.94 (0.75–1.18) Social support: ♂:69/1377, 1.0 (0.75–1.33), ♀: 31/585, 1.08 (0.70–1.67) All: 100/1962, 1.02 (0.81–1.30) Job strain: ♂: 43/987, 0.82 (0.59–1.15) ♀: 35/512, 1.59 (1.03–2.45) All: 78/1499, 1.04 (0.80–1.34) Iso–strain: ♂: 25/475, 1.07 (0.71–1.63) ♀: 20/24, 1.94 (1.17–3.21) All: 45/716, 1.33 (0.97–1.83) |
| Heraclides et al, 2012 (39) United Kingdom (Whitehall II) |
Baseline: 1991–1993 FU: 18y Workers: 5138 /10308 PB: 73%, PI 72% 1449/3689 Mean age: 49y |
Exposure at baseline JCQ Demands: 4 items Job control: 15 items Job strain: ⫿25%, ⫿32%, all 27% |
OGTT level and self-report 927 cases (18%) | Cox regression model, HR (95% CI) PCE: Yes Stratified by sex, controlled by age, employment grade, diet pattern, alcohol consumption, physical activity, smoking status, systolic blood pressure, triglycerides, high–density lipoprotein cholesterol ♂: 389/3.689, high strain: 0.80 (0.63–1.02) ♀: 151/1.449, high strain: 1.37 (0.98–1.92) |
| Hino et al, 2016 (32) Kanto Japan |
Baseline: 2008 and 2011 FU: 3y Male workers: 1815 / 29 586 PB: 21%, PI 43% Age: ≥35y |
2 waves exposure: BJSQ Demands: 3 items Job control: 3 items Social support: 3 items each colleagues resp. supervisor: Demand increase: 9% Control increase: 14% SS supervisor incr.: 12% Coll. support incr.: 12% Job strain incr.: 9% |
FPG, HbA1c, and immuno-reactive insulin (IRI) levels (≥2.5 on HOMA-IR) 136 cases (7.5%) | Logistic regression, OR (95% CI) PCE: Yes Age, marital status, occupational characteristics (job department, employment position and occupation Demands increase: 0.45 (0.19–1.03) Job control increase: 0.59 (0.31–1.12) Colleagues support increase: 0.86 (0.48–1.56) Supervisor support increase: 0.65 (0.33–1.28) Job strain increase: 0.56 (0.25–1.22) |
| Huth et al, 2014 (30) Germany (MONIKA/KORA) |
Baseline: 1984-1994 FU: 12,7y Workers: 5337 /17 438 PB: 75%, PI 73% 1986 / 3351 Mean age: 43.9y |
Exposure at baseline Adaptation of JCQ Demands: 5 items Job control: 6 items High strain: 19% |
Self-reported and the date of diagnosis validated by hospital 291 cases (5.5%) | Cox proportional hazard model, HR (95% CI) PCE: Yes Age, sex, baseline survey, education and physical intensity work Job strain: 1.24 (0.93–1.65) High vs. low strain: 1.43 (1.00–2.06) |
| Kawakami et al, 1999 (35) Japan (Electrical) |
Baseline: 1984 FU: 8y Male workers: 2194 / 3862 PB:92%, PI:77% Range: 18-60 y |
Exposure at baseline Adaptation of JCQ Demands: 1 question Job control: 1 question Support: 1 question Job strain: 10% Low support: 19% |
FPG ≥110 mg/dl + oral glucose tolerance test 34 cases (1,5%) | Cox proportional hazard model, HR (95% CI) PCE: Yes Age, education, BMI, alcohol, consumption, smoking, leisure time, physical activity, family history Job strain: 1.34 (0.50–3.55) Low social support: 1.27 (0.58–2.79) |
| Kroenke et al, 2007 (12) USA (NHS II) |
Baseline: 1993, FU 6y Women nurses: 62 574 / 116 608 PB:75%, PI: 73% Mean age: 38.8y Range: 29-46y |
Exposure at baseline JCQ, 27 items High strain: 20% |
Self-reported, high confirmation rate, 365 cases (5.8%) | Cox proportional hazard model, HR (95% CI) PCE: Yes Age High strain: 1.13 (0.84–1.51) |
| Kumari et al, 2004 (36) United Kingdom (Whitehall II) |
Baseline: 1992-93, FU 5-6y White-collar workers: 8386 / 10 308 PB:73%, PI 82% 2579 / 5807 |
Exposure at baseline JCQ Demands: 4 questions Job control: 15 questions Support: 6 questions ERI (proxy, no information on the number of items per dimension) high effort, low reward Fraction NI |
OGTT level and self-report 361 cases (4.3%) | Logistic regression, OR (95% CI) PCE: Yes age, length of follow–up, employment grade, ethnic group and ECG abnormalities Demands: ♂ 1.11 (07–1.7), ♀: 0.59 (0.3–1.2) Job control: ♂ 0,77 (0.5–1.2), ♀: 0.82 (0.4–1.6) Support: ♂ 0.80 (0.5–1.1), ♀: 1.20 (0.7–1.9) ERI highest vs lowest category: ♂ 1.71 (1.0–2.8), ♀: 0.92 (0.4–1.9) |
| Mortensen et al, 2017 (29) France (GAZEL) Sweden (SLOSH) United Kingdom (Whitehall II) |
Baseline: 2000 (GAZEL), 2006 (SLOSH), 1991–1994 (Whitehall II) FU 10 y White-collar workers: GAZEL: 6572/ 20625 SLOSH: 7590/40877 Whitehall: 7081/10308 PB: 45% (GAZEL), ~39% (SLOSH), 73% (Whitehall II) PI: 51% (GAZEL), 46% (SLOSH), 77% (Whitehall II) 8710 / 12 533 Range: 46–55 y |
Exposure at baseline JCQ Demands:5 items Job control: 6 items Support: 2 items High strain 22% Low support 36% |
Self-reported complemented with OGTT and FPG 1058 cases (433; 208; 417 resp) (5.0%) | Logistic regression OR (95% CI) PCE: Yes Age, sex, marital status, occupation and sub–cohort High job strain: GAZEL: 1.24 (0.93–1.64) SLOSH: 0.96 (0.65–1.41) Whitehall: 0.93 (0.70–1.22) Low support: GAZEL: 1.07 (0.85–1.35) SLOSH: 1.26 (0.93–1.69) Whitehall: 1.27 (0.99–1.64) |
| Mutambudzi et al, 2016 (40) USA (HRS) |
Baseline: 2006 FU: 7y Middle- and older-aged workers: 1396 / 18 469 PB: 74%, PI:19– 50% Mean age: 58 y range ≥50 |
Exposure at baseline JCQ-like Demands: 3 items Job control: 3 items High strain 11% |
Self-reported 167 cases (11.5%) | Cox proportional hazard model, HR (95% CI) PCE: Yes Adjusted for body mass index, physical activity, education, race, gender, alcohol use, average work hours/week, occupational category, marital status, insurance coverage, and hypertension. Low strain was treated as the referent category. High strain vs low strain: 1.73 (1.09–2.75) |
| Mutambudzi et al, 2018 (37) USA (HRS) |
Baseline: 2006 FU: 7y Middle- and older-aged workers 1932 / 18 469 PB: 74%, PI: 24–59% 1041 / 894 Mean age: 61 y range ≥50 |
Exposure at baseline ERI Effort: 2 items Reward: 5 items ERI 25% |
Self-reported 288 cases (11.8%) | Cox proportional hazard model, HR (95% CI) PCE: Unclear Age, sex, race, education, marital status ERI: 1.18 (0.94–1.48) |
| Norberg et al, 2007 (33) Sweden (VIP) |
Baseline: 1989-2000, FU: 12y Nested case-cohort workers 191 cases, 393 controls PB 52%, PI: NA 240/344 Range: 40 or 50 or 60 y at baseline |
Exposure at baseline JCQ Demands+ Job control: 10 items High strain: 11% |
Administrative data 191 cases | Logistic regression, OR (95% CI) PCE: Yes Matched by age, sex and survey year High strain vs low strain: ♂ 1.00 (0.5–2.00), ♀ 2.8 (1.1–7.6) |
| Nordentoft et al, 2020 (20) Denmark (WEHD) |
Baseline: 2012, 2014, 2016 FU 2.7 y Active general population: 50 552 / 115 564 PB: 54%, PI 97% 26 378 / 26 378 Range: 30-64 y |
Exposure at baseline ERI Effort: 6 items Reward: 5 items ERI 25% |
Administrative data 347 cases (0.69%) | Cox proportional hazard model, HR (95% CI) PCE: Unclear sex, age, cohabitation, young children in the household, SES, migration background, survey year and sample method Dichotomic ERI: both sexes 1.27 (1.02–1.58) Continuous ERI: both sexes 1.09 (0.98–1.21), ♂ 1.09 (0.95–1.25), ♀ 1.08 (0.93–1.26) |
| Nyberg et al, 2014 (11) Europe (IPD-Work) |
Baseline: 1986-2008, 13 individual studies, FU: 23y Workers: 124 808 / NI PB: 41-82%, PI: 53-98% 70802/54006 Mean age: 49 y |
Exposure at baseline JCQ harmonized Demands: 2-6 items Job control: 5-6 items Job strain 16% |
Depending on the individual studies: Self-reported, OGTT and administrative records 3703 cases (3.0%) |
Cox proportional hazard model, HR (95% CI) PCE: Yes Age, sex, occupational title Dichotomous job strain: 1.15 (1.06–1.25) High strain vs. low strain (supplementary results published separately in (25)): 1.13 (1.02–1.25) |
| Pan et al, 2017 (16) Sweden (SNAC-K) |
Baseline: 2001–2004, FU: 3 y for >78 years and 6y for <78 years old Retired workers: 2719 / 3363 PB: 73%, PI: 88% 1756/963 Mean age: 49 y |
Exposure during work life JCQ-Matrix Demands: 2 items Job control: 12 items Median score for each type of occupation Job strain 21% |
Self-reported or administrative data or HbA1c >6.4% 154 cases (5.7%) | Logistic regression, OR (95% CI) PCE: Yes Age, sex, educational level, follow–up time Job strain: 1.60 (1.07–2.39) |
| Smith et al, 2012 (31) Canada (CCHS) |
Baseline: 2000-2001, FU 10y Workers 7443/NI PB: 84% NI, PI: 89.6% 3752/3691 range: 35-60 |
Exposure at baseline JCQ-like Demands: 2 items Job control: 5 items Support: 3 items Low demands: ♂ 23%, ♀ 18%. Low job control: ♂ 20%, ♀26%. Low support: ♂18%, ♀18% |
Administrative data 639 cases (8.7%) | Cox proportional hazard model, HR (95% CI) PCE: Yes Age, immigration status, ethnicity, marital status, urban or rural living location, education, heart disease at baseline, hypertension at baseline, depression at baseline, activity limitations at work due to health problems, and other work variables (shift schedule, weeks worked, multiple jobs, physical activity at work). Low demands (4th quartile): ♂: 0,72 (0.45–1.14), ♀ 0.75 (0.43–1.33), Low control (4th quartile): ♂: 0,84 (0.48–1.45), ♀ 2.17 (1.23–3.83) Low support (4thquartile): ♂:1.19 (0.68–2.10), ♀ 0.43 (0.23–0.82) |
| Souza Santos et al, 2020 (19) Brazil (Elsa-Brasil) |
Baseline:2008-2010, FU y Workers civil servants 7503/ 52137 PB: 29%, PI: 86% 3998/3505 Mean age: 52y range: 35-74 |
Exposure at baseline DCS, ERI Demands: 5 items Job control: 6 items Support: 6 items Effort: 6 items Reward: 10 items All measures for tertiles. High job strain: ♀39.1% ♂ 28.8% Low SS: ♀36.2% ♂ 30.4% Iso-strain: ♀24.9% ♂ 16.7% High effort/reward: ♀ 37.1♂ 32.6% High Overcommitment: ♀ 31.9 ♂ 29.6% High Job strain + ERI: ♀25.7% ♂ 16.7% |
HbA1c ≥6.5% 167 cases (2.2%) | Logistic regression, OR (95% CI) PCE: yes Age, schooling level, weekly workload, work shift. Most unfavorable tertile vs. most favorable tertile. High demands: ♂1.22 (0.66–2.27), ♀2.41 (1.30–4.50) Low job control: ♂0.95 (0.51–1.77), ♀0.64 (0.34–1.18) High job strain: ♂1.02 (0.59–1.76), ♀1.77 (0.98–3.19) Low support: ♂1.29 (0.69–2.40), ♀1.93 (0.96–3.87) High effort: ♂1.14 0.64–2.04), ♀1.17 (0.68–2.01) Low reward: ♂0.96 (0.55–1.68), ♀1.76 (1.03–2.99) High effort/reward: ♂1.08 (0.62–1.88), ♀1.36 (0.81–2.29) High Overcommitment: ♂ 0.81(0.44–1.49), ♀ 1.46 (0.80–2.63) High Job strain + ERI: ♂ 0.92 (0.49–1.75), ♀ 2.10 (1.20–3.65) |
| Toker et al, 2012 (43) Israel |
Baseline: 2003 and 2008 FU: 8y 5843 / 12754 PB: 92%, PI 55% NI Mean age: 48 y |
Exposure at baseline JCQ-like Demands: 6 items Job control: 7 items Social support: 8 items Fractions NI |
FPG ≥ 126 mg/dL or HbA1c ≥ 6.5% or Self-reported 182 cases (3,1%) |
Logistic regression, OR (95% CI) PCE: Yes Age, sex, education, follow–up time, family history, LDL, body mass index, systolic pression, triglycerides, smoking, physical activity, depression High demands: 0.98 (0.83–1.15) Low job control: 1.05 (0.85–1.29) Low support: 0.79 (0.62–0.99) |
| Yamaguchi et al, 2018 (41) Japan (Furukawa Nutrition and Health Study) |
Baseline: 2012-2013 FU: 3 y 1040/ 2828 PB: 76%, PI 56% 115/925 Mean age: 42 y Range:19-68 |
Repeated exposure in two waves JCQ-like Demands: 5 items Job control: 9 items Social support: 8 items Job strain increase: 16% |
FPG ≥100 mg/dL 64 cases (6.8%) | Logistic regression, OR (95% CI) PCE: Unclear, exclusion of subjects with metabolic syndrome age, sex, site, family structure, marital status, occupational category, work status, night or rotating shift work, work–related physical activity, leisure–time physical activity, smoking, alcohol drinking, sleep duration, quality of sleep, energy intake, and each component of metabolic syndrome at baseline. Job strain increase: 3.86 (1.77–8.38) |