Table 1.
Characteristic of included epidemiological studies.
| Author | Study type | Population | Country | No. of subjects | Specific air pollutant | Outcome | Main findingsa | Exposure duration |
|---|---|---|---|---|---|---|---|---|
| Effect of air pollution exposure duration and dose–response on T2DM | ||||||||
| Hassanvand9 | Cross-sectional study | Iranian adults | Iran | 2916 | PM10 | OR | T2DM [1.32 (95% CI 1.03, 1.69)] | Long-term (6 years) |
| Qiu10 | Cohort study | Aged 65+ | Hong Kong | 61,447 | PM2.5 | OR, HR | Prevalence of T2DM [PM2.5 1.06 (95% CI 1.01, 1.11)]; incidence of T2DM [PM2.5 1.15 (95% CI 1.05, 1.25)] | Long-term (1 year) |
| Andersen11 | Cohort study | Aged 50–65 | Denmark | 57,053 | NO2 | HR | T2DM [1.04 (95% CI 1.00, 1.08)] | Long-term (9.7 year) |
| Wolf12 | Cross-sectional study | German adults | Germany | 3080 | PM10 | Beta coefficient | HOMA-IR [15.6% (95% CI 4.0, 28.6)] HbA1c [0.5% (95% CI −0.5, 1.5)] | Long-term (3 years) |
| Lucht13 | Cohort study | Nondiabetic patients | Germany | 7108 | PM2.5 | Beta coefficient | FBG [28-day 0.91 (95% CI 0.38, 1.44) 91-day 0.81 (95% CI 0.05, 1.58)] HbA1c [28-day 0.03 (95% CI 0.01–0.05); 91-day 0.07 (95% CI 0.04, 0.10)] | Medium-term (28–91 days) |
| Li14 | Cross-sectional study | T2DM patients | China | 2840 | PM2.5, PM10, SO2, CO | Beta coefficient | LOS [PM2.5 (+) PM10 (+) SO2 + CO +]; hospitalization costs [PM2.5 (–) PM10 (–) SO2 + CO +] | Short-term (up to 16 days) |
| Lucht15 | Cohort study | Aged 45–75 | Germany | 6727 | PM2.5, PM10, NO2 | Beta coefficient | Adiponectin [PM2.5 (–) PM10 − NO2 −] IL-1RA; (PM2.5 + PM10 + NO2 +) | 1–365 days |
| Kim16 | Cohort study | Aged 60+ | Korea | 1560 | PM10, NO2, O3, SO2 | Beta coefficient | IR [PM10 + NO2 + O3 + SO2 (+)] | Short-term (up to 10 days) |
| Bowe17 | Cohort study | US veterans without diabetes | United States | 172,9108 | PM2.5 | Response function | Incidence of T2DM + | Long-term (more than 1 year) |
| Honda18 | Cross-sectional study | Aged 57+ | United States | 4121 | PM2.5, NO2 | Prevalence; OR | T2DM [PM2.5 1.35 (95% CI 1.19, 1.53); NO2 1.27 (1.10, 1.48)] | Long-term (more than 1 year) |
| Tan19 | Cross-sectional study | Policemen | China | 1271 | PM2.5 | Beta coefficient | FBG [0.005% (95% CI 0.0004%, 0.009%)] | Long-term (more than 1 year) |
| Effects of air pollution exposure on T2DM in subpopulations | ||||||||
| Qiu10 | Cohort study | Aged 65+ | Hong Kong | 61,447 | PM2.5 | OR, HR | Prevalence of T2DM [PM2.5 1.06 (95% CI 1.01, 1.11)]; incidence of T2DM [PM2.5 1.15 (95% CI 1.05, 1.25)] | Long-term (1 year) |
| Honda18 | Cohort study | Aged 57+ | United States | 4121 | PM2.5, NO2 | Beta coefficient, OR | HbA1c for diabetic (PM2.5 + NO2 +); HbA1c for nondiabetic [PM2.5 (+) NO2 +]; T2DM [PM2.5 1.35 (95% CI 1.19, 1.53); NO2 1.27 (1.10, 1.48)] | Long-term (more than 1 year) |
| Yang20 | Cross-sectional study | Aged 50+ | China | 11,504 | PM2.5 | OR | T2DM [1.27 (95% CI 1.12, 1.43)] | Long-term (3 years) |
| Shen21 | Case-control study | Female | Taiwan | 6717 | PM2.5, SO2 | OR | GDM [PM2.5 1.10 (95% CI 1.03, 1.18); SO2 1.37 (95% CI 1.30, 1.45)] | Medium-term (12 weeks) |
| van den Hooven22 | Cohort study | Pregnant women | Netherlands | 7339 | Traffic density | Beta coefficient | GDM (0) | Long-term (more than 1 year) |
| Fleisch23 | Cohort study | Pregnant women | United States | 2093 | PM2.5 | OR | IGT [2.63 (95% CI 1.15, 6.01)]; GDM [0.71 (95% CI 0.35, 1.42)] | Long-term (more than 1 year) |
| Hansen24 | Cohort study | Female nurses | Denmark | 24,174 | PM2.5 PM10 NO2 NOX | HR | T2DM [PM2.5 1.11 (95% CI 1.02, 1.22); PM10 1.06 (95% CI 0.98, 1.14); NO2 1.05 (95% CI 0.99, 1.12); NOX 1.01 (95% CI 0.98, 1.05)] | Long-term (5 years) |
| Eze25 | Meta-analysis | N/A | Europe and North America | N/A | PM2.5, NO2 | OR | T2DM [PM2.5 1.10 (95% CI 1.02, 1.18); NO2 1.08 (95% CI 1.00, 1.17)] | Long-term (more than 1 year) |
| Lim26 | Cohort study | Aged 50+ | United States | 549,735 | PM2.5 NO2 03 | HR | Diabetes mortality [PM2.5 1.19 (95% CI 1.03, 1.39); NO2 1.09 (95% CI 1.01, 1.18); O3 0.96 (95% CI 0.88, 1.04)] | Long-term (30 years) |
| Effect of air pollution exposure on T2DM in countries on different economic development levels | ||||||||
| Bai27 | Cohort study | Aged 30+ | Canada | 105,6012 | NO2 | HR | T2DM [1.06 (95% CI 1.05, 1.07)] | Long-term (more than 1 year) |
| Hernandez28 | Cross-sectional study | Adults | United States | 862,519 | PM2.5, O3 | Prevalence ratio | T2DM (PM2.5 1.10 (95% CI 1.03, 1.17); O3 1.06 (95% CI 1.03, 1.09)] | Long-term (7 years) |
| Eze29 | Cross-sectional study | Adults | Swiss | 6392 | PM10, NO2 | OR | T2DM [PM10 1.40 (95% CI 1.17, 1.67); NO2 1.19 (95% CI 1.03, 1.38)] | Long-term (10 years) |
| Renzi30 | Cohort study | Aged 35+ | Italy | 65,955 | NO2, NOX, O3 | OR, HR | Prevalence of T2DM [NO2 1.010 (95% CI 1.002, 1.017); NOX 1.015 (95% CI 1.009, 1.021)]; incidence of T2DM [NOX 1.011 (95% CI 1.003, 1.019); O3 1.015 (95% CI 1.002, 1.027)] | Long-term (3 years) |
| Kramer8 | Cohort study | Nondiabetic women | Germany | 1775 | PM10, NO2 | HR | T2DM [PM10 1.16 (95% CI 0.81, 1.65); NO2 1.34 (95% CI 1.02, 1.76)] | Long-term (5 years) |
| Coogan31 | Cohort study | Black women | United States | 4204 | PM2.5 | Incidence rate ratios | T2DM [1.63 (95% CI 0.78, 3.44)] | Long-term (1 year) |
| Puett32 | Cohort study | Female nurses | United States | 74,412 | PM2.5, PM10 | HR | T2DM [PM2.5 1.02 (95% CI 0.94, 1.09); PM10 1.03 (95% CI 0.98, 1.09)] | Long-term (1 year) |
| Dijkema33 | Cross-sectional study | Aged 50–75 | Netherlands | 8018 | NO2 | OR | T2DM [0.80 (95% CI 0.63, 1.02)] | Long-term (2 years) |
| Yang20 | Cross-sectional study | Aged 50+ | China | 11,504 | PM2.5 | OR | T2DM [1.27 (95% CI 1.12, 1.43)] | Long-term (3 years) |
| Yang6 | Cross-sectional study | Adults | China | 15,477 | PM1, PM2.5, PM10 | OR | T2DM [PM1 1.13 (95% 1.04, 1.22); PM2.5 1.14 (95% CI 1.03, 1.25); PM10 1.20 (95% CI 1.12, 1.28)] | Long-term (3 years) |
| Effects of different air pollution components | ||||||||
| Weinmayr34 | Cohort study | Individuals without diabetes | Germany | 3607 | PM2.5 PM10 | Relative risk | T2DM [PM2.5 1.03 (95% CI 0.95, 1.12); PM10 1.05 (95% CI 1.00, 1.10)] | Long-term (1 year) |
| Strak35 | Cross-sectional study | Adults | Netherlands | 289,703 | PM2.5 PM10 NO2 | OR | T2DM [PM2.5 1.01 (95% CI 0.99, 1.03); PM10 1.04 (95% CI 1.02, 1.06); NO2 1.07 (95% CI 1.05, 1.09)] | Long-term (more than 1 year) |
| Sun36 | Cross-sectional study | Aged 65+ | Hong Kong | 40,150 | Elemental carbon, organic carbon, nitrate, and nickel | Excess risk | T2DM admissions [elemental carbon 3.79% (95% CI 1.63, 5.95); organic carbon 3.74% (95% CI 0.83, 6.64); nitrate 4.58% (95% CI 2.17, 6.99); nickel 1.91% (95% CI 0.43, 3.38)] | Short-term (24 h) |
| Zanobetti37 | Cross-sectional study | Adults | United States | 46,192 | Elemental carbon, organic carbon, As, SO42− | Excess risk | T2DM admissions [organic carbon −2.42% (95% CI −3.79, −1.06); elemental carbon −2.12% (95% CI −3.84, −0.39); As 2.16% (95% CI 0.11, 4.21); SO42− 2.91% (95% CI 0.92, 4.89)] | Short-term (48 h) |
a
The direction of the effect is denoted by − for significant negative association; (–) for negative association; (0) for no association; + for significant positive association; (+) for positive association.
As, arsenic; CI, confidence interval; CO, carbon monoxide; FBG, fasting blood glucose; GDM, gestational diabetes mellitus; Hb1Ac, hemoglobin 1Ac; HOMA-IR, homeostatic model assessment of insulin resistance; HR, hazard ratio; IGT, impaired glucose tolerance; IL-IRA, interleukin 1RA; IR, insulin resistance; LOS, length of stay; NO2 nitrogen dioxide; NOX, nitrogen oxides; O3, ozone; OR, odds ratio; PM2.5, fine particulate matter; PM10, inhalable particles; SO2 sulfur dioxide; T2DM, type 2 diabetes mellitus.