Skip to main content
NCBI home page
Elsevier Sponsored Documents logoLink to Elsevier Sponsored Documents
. 2023 Nov 1;897:165235. doi: 10.1016/j.scitotenv.2023.165235

The effects of long-term exposure to air pollution on incident mental disorders among patients with prediabetes and diabetes: Findings from a large prospective cohort

Jin Feng a, Miao Cai a, Zhengmin (Min) Qian b, Shiyu Zhang a, Yin Yang a, Stephen Edward McMillin c, Ge Chen a, Junjie Hua a, Maya Tabet d, Chongjian Wang e, Xiaojie Wang a, Hualiang Lin a,⁎
PMCID: PMC10522921  PMID: 37414192

Abstract

Background

The association between air pollution and mental disorders has been widely documented in the general population. However, the evidence among susceptible populations, such as individuals with prediabetes or diabetes, is still insufficient.

Methods

We analyzed data from 48,515 participants with prediabetes and 24,393 participants with diabetes from the UK Biobank. Annual pollution data were collected for fine particulate matter (PM2.5), inhalable particulate matter (PM10), nitrogen dioxide (NO2), and nitrogen dioxides (NOx) during 2006–2021. The exposure to air pollution and temperature for each participant were estimated by the bilinear interpolation approach and time-weighted method based on their geocoded home addresses and time spent at each address. We employed the generalized propensity score model based on the generalized estimating equation and the time-varying covariates Cox model to assess the effects of air pollution.

Results

We observed causal links between air pollutants and mental disorders among both prediabetic and diabetic participants, with stronger effects among those with diabetes than prediabetes. The hazard ratios were 1.18 (1.12, 1.24), 1.15 (1.10, 1.20), 1.18 (1.13, 1.23), and 1.15 (1.11, 1.19) in patients with prediabetes, and 1.21 (1.13, 1.29), 1.17 (1.11, 1.24), 1.19 (1.13, 1.25), and 1.17 (1.12, 1.23) in patients with diabetes per interquartile range elevation in PM2.5, PM10, NO2, and NOx. Furthermore, the effects were more pronounced among people who were older, alcohol drinkers, and living in urban areas.

Conclusions

Our study indicates the potential causal links between long-term exposure to air pollution and incident mental disorders among those with prediabetes and diabetes. Reducing air pollution levels would significantly benefit this vulnerable population by reducing the incidence of mental disorders.

Keywords: Air pollution, Mental disorders, Prediabetes, Diabetes, Causal inference

Graphical abstract

Unlabelled Image

Open in a new tab

Highlights

  • •

    People with prediabetes and diabetes were susceptible to long-term air pollution.

  • •

    Long-term air pollution had adverse effect on incident mental disorders.

  • •

    Patients who were older, alcohol drinkers, and urban residence were more vulnerable.

1. Introduction

Diabetes is a significant and pressing global threat to human health in the twenty-first century. According to the International Diabetes Federation, approximately 541 and 537 million adults aged 20–75 years are suffering from prediabetes and diabetes worldwide in 2021. Based on the 2019 Global Burden of Disease, the age-standardized disability-adjusted life-year rate of type 2 diabetes all over the world was 801.55 per 100,000 (Kotwas et al., 2021; Safiri et al., 2022). In addition, mental disorders are also a major challenge to human health, contributing to 4.9 % of global disability-adjusted life-years and 14.6 % of years lived with disability in 2019 (Arias et al., 2022; Collaborators, 2022).

Compared to the general population, people with diabetes are more prone to mental disorders (Boden, 2018; Garrett and Doherty, 2014). A systematic review found that the prevalence of depression in people with diabetes was almost 2–3 times higher than that in people without diabetes (Roy and Lloyd, 2012). Additionally, some studies indicated that individuals with diabetes were more likely to develop anxiety disorders (Buchberger et al., 2016; Cobham et al., 2020; Majidi et al., 2015). Abnormal blood glucose may lead to changes in brain function and metabolism including hippocampal atrophy, prefrontal cortex dysfunction, and neurotransmitter imbalance (Fu et al., 2022; Zhang et al., 2021). Moreover, strict self-management requirements for diet and exercise may also lead to abnormalities in diabetic patients' emotional and cognitive functions. (Benton et al., 2023). The coexistence of diabetes and mental illness can complicate treatment and elevate the risk of adverse outcomes (Baumeister et al., 2011; Egede et al., 2005). Therefore, it is of great importance to identify modifiable risk factors among people with prediabetes and diabetes to prevent the incidence of mental disorders.

Air pollution is associated with mental disorders (Li et al., 2020; Ventriglio et al., 2021; Yang et al., 2023). However, findings have mainly come from studies on the general population rather than people with a medical condition. Only one ecological study has investigated the association between air pollutants and emergency department visits for depressive episodes in patients with diabetes mellitus, but similar individual-level studies are still lacking (Cho et al., 2014). Furthermore, current findings are typically based on traditional approaches, which may be biased due to the unbalanced distribution of covariates. Therefore, we conducted the present study to examine the association between long-term exposure to air pollutants and incident mental disorders among people with prediabetes and diabetes using the causal modeling method.

2. Methods

2.1. Study design and participants

UK Biobank is a large prospective cohort of 502,490 adults aged 40–69 years in England, Scotland, and Wales. Details of the study design and the data collection have been described elsewhere (Caleyachetty et al., 2021). Briefly, the baseline survey was conducted during 2006–2010, which collected the demographic and socio-economic status of the participants, as well as health-related information through a questionnaire and physical measurements. UK Biobank study protocol was approved by the North West Multicenter Research Ethics Committee. Written informed consents were obtained from each participant (Zhang et al., 2022a).

The definition of diabetes included: 1) Glycated hemoglobin (HbA1c) ≥6.5 % [48 mmol/mol] (Cai et al., 2021); 2) Diagnosed diabetes by a physician; 3) Diagnosed diabetes by ICD10 code when admitted to the hospital (Atkins et al., 2020); 4) Self-reported diabetes; 5) Medication history of insulin or hypoglycemic medication. Those who were not classified as having diabetes but had HbA1c between 5.7 % [39 mmol/mol] and 6.5 % [48 mmol/mol] were defined as having prediabetes (Zhang et al., 2022a).

In this study, we excluded participants with incomplete data on HbA1c and air pollution. We further excluded participants without prediabetes or diabetes and participants with a medical history of mental disorders at baseline. In total, 48,515 people with prediabetes and 24,393 with diabetes were included in the analyses (Fig. S1). We conducted multiple imputations with chained equations to impute missing data on covariates. The fundamental characteristics of imputed and non-imputed participants were shown in Table S8.

2.2. Air pollution exposure

We extracted grid data of annual average concentrations of fine particulate matter (PM2.5), inhalable particulate matter (PM10), nitrogen dioxide (NO2), and nitrogen dioxides (NOx) during 2006–2021 from the UK AIR dataset, which provides annual concentration maps for various pollutants at 1 km × 1 km spatial resolution (UK Air, 2022). These maps were calculated by summing contributions from different sources of the National Atmospheric Emissions Inventory, a combination of air-pollutant and greenhouse-gas data (2022). The bilinear interpolation approach was used to estimate annual exposure to air pollution for each participant based on their geocoded home addresses (Cai et al., 2023).

In addition, considering each patient's home moving history, we used a time-weighted method to estimate exposure more accurately. Specifically, we first obtained the patient's home addresses and the length of time they had lived at each address. We further used the time spent at each address as a weight to calculate the average exposure within each year. More detailed information about exposure assessment was shown in the previous study (Wu et al., 2022).

2.3. Temperature exposure

The monthly ambient temperature with 1 km × 1 km spatial resolution was extracted from the HadUK-Grid dataset, which was produced by the Met Office Hadley Centre for Climate Science and Services (Dan Hollis et al., 2019). This dataset was created by interpolating station observations into a regular grid using inverse-distance-weighted interpolation of residuals from a multiple regression model (Perry and Hollis, 2005). Similar to the exposure assessment of air pollution, the bilinear interpolation approach combined with the time-weighted method was applied to estimate annual meteorological exposure for each individual.

2.4. Outcomes

All participants were followed up until the first occurrence of mental disorders, loss to follow-up, death, or January 1st, 2022 (the latest date available for the air pollution data), whichever came first.

The main outcomes of this study were overall mental disorders, depressive disorders, and anxiety disorders. These outcomes were identified via linkage to the participants' primary care, hospital admission, and the UK National Health Service. We used International Classification of Disease Edition 10 (ICD-10) codes F00–99, F32–33, and F40–41 to define overall mental disorders, depressive disorders, and anxiety disorders, respectively (Pan et al., 2022).

2.5. Covariates

In accordance with the literature (Hao et al., 2022; Roberts et al., 2019), we considered the following covariates in our analyses: age, sex (female or male), ethnicity (White or non-White), household income (<18,000£, 18,000-30,999£, 31,000-51,999£, ≥52,000£, or unknown), educational level (college or university degree, any school degree, vocational qualifications or other), body mass index (underweight (<18.5 kg/m2), normal weight (18.5–25 kg/m2), overweight (25–30 kg/m2), obsess (≥30 kg/m2)), smoking status (never, former, or current), alcohol intake (never, occasional, moderate, or heavy), physical activity (low, moderate, high, or unknown; assessed by the International Diabetes Federation, 2021), Townsend deprivation index (low, moderate, high) (Yang et al., 2022), time spent outdoors (low, moderate, high), residential area (rural or urban; determined by the population density of the participants' postcodes) (Cai et al., 2022), percentage of greenspace (%) at 300 m buffers (low, moderate, high) (Zhang et al., 2022b), insulin or hypoglycemic medication (yes or no), and number of comorbidity diseases (including hypertension, stroke, heart failure, coronary heart disease, chronic obstructive pulmonary disease, chronic bronchitis, migraine, arthritis, and malignant tumor). Detailed descriptions of these variables are provided in Table S1.

We further plotted a directed acyclic graph (DAG) incorporating all the above variables to select confounding variables. Based on the DAG, we controlled for age, sex, ethnicity, household income, physical activity, Townsend deprivation index, time spent outdoors, residential area, percentage of green space, and ambient temperature (Fig. S2).

2.6. Statistical analyses

We created a counting process dataset for cohort survival data (Andersen and Gill, 1982). Each participant was followed up from the recruitment date to the failure-event date. Each row of the dataset represented a person followed for one year (Heine-Bröring et al., 2013). We assumed that annual air pollution, ambient temperature, and age were time-varying exposures, which varied from year to year. Time-varying covariates Cox model and the generalized propensity score model were applied to estimate the association between exposure to air pollution and mental disorders among people with prediabetes and diabetes.

2.6.1. Generalized propensity score model

The generalized propensity score (GPS) method was the extension of the propensity score (PS) approach to continuous exposures, which was initially developed by Keisuke Hirano and Guido W. Imbens (Imbens, 2000; Keisuke Hirano and Imbens, 2004). The generalized propensity score is a balancing score based on the conditional density of the continuous exposure given the confounders (Austin, 2019; Gao et al., 2021). We estimate GPSs by regressing the exposure to air pollutants on the observed covariates using a generalized estimating equation (GEE) (Wang et al., 2023). Stabilized inverse probability weights (IPW) were calculated by the inverse of GPSs using the following formula:

Stabilized Weight=fxfx/v

In the model, x represents the exposure and v represents the covariates including age at recruitment, sex, ethnicity, household income, physical activity, Townsend deprivation index, time spent outdoors, residential area, the percentage of green space, and ambient temperature. The f (x) is the probability density function of the exposure. The f (x/v) is the probability density function of the exposure given the covariates, which are GPSs. We truncated the lowest and highest 1 % of the IPW weights to stabilize the estimations (Wang et al., 2017). We further calculated averaged absolute correlation value (AC) to test covariate balance after weighting. Averaged absolute correlation value lower than 0.1 was considered as good covariate balance (Wu et al., 2020).

2.6.2. Time-varying covariates cox model

We fit non-weighted and IPW-weighted time-varying covariates Cox models to calculate hazard ratios (HRs) and 95 % confidence intervals (CIs) (Xu et al., 2014). In this model, we adjusted for the covariates specified by DAG. We further included the calendar year as a categorical variable to control the long-time trend (Wu et al., 2020). The model is specified as follows:

Survivalstart datestop datestatus∼IQRAirpollutant+Age+Sex+Ethnicity+Household income+Physical activity+Time spent outdoors+Greenspace+Residential area+Townsend deprivation index+Ambient temperature+strataYear

2.6.3. Stratified analyses

We conducted stratified analyses by sex, age, educational level, residential area, smoking status, and alcohol intake. All the covariates were included except for the strata variables. We further included the product term of air pollution and stratification factors to identify the potential interactions (Nissen et al., 2022).

2.6.4. Sensitivity analyses

We conducted several sensitivity analyses to examine the robustness: (1) calculating the E-value. E-value is defined as the minimum strength of association on the risk ratio scale that an unmeasured confounder would need to have with both the exposure and the outcome, conditional on the measured covariates, to fully explain away a specific exposure-outcome association (VanderWeele and Ding, 2017). E-value was used to measure the magnitude of the unmeasured confounding. A higher E-value suggests less unmeasured confounding in the analyses (Ioannidis et al., 2019); (2) adding a cubic spline for continuous variables; (3) excluding participants with mental disorder events or death within 1 year of follow-up to control for competing diseases; (4) excluding the participants with missing data on covariates; (5) conducting 1:1 propensity score matching with 0.1 calipers between participants with prediabetes and diabetes; (6) conducting the analyses by marginal structural models (Wang et al., 2023); (7) conducting the analyses by Cox proportional-hazards model with exposure at baseline; (8) calculating crude hazard ratios; (9) restricting the analyses to participants with air-pollution concentrations generally below the third quartile of the entire cohort throughout the study period (Wang et al., 2023).

All the analyses were performed using R software (version 4.1.3). A two side P<0.05 was considered statistically significant.

3. Results

3.1. Descriptive results

A total of 48,515 participants with prediabetes and 24,393 participants with diabetes at baseline were included in this study (Table 1). Among people with prediabetes or diabetes, 36,587 (50.2 %) were men, the mean (SD) age was 59.6 (7.2) years, and 64,833 (88.9 %) were white. The annual mean (SD) temperature was 9.8 (2.4) °C. Moreover, the annual concentration of air pollutants including PM2.5, PM10, NO2, and NOx at baseline were higher in mental-disorder cases [PM2.5: 9.8 (SD 1.7) μg/m3; PM10: 14.6 (SD 2.4) μg/m3; NO2: 17.8 (SD 6.0) μg/m3; NOx: 26.5 (SD 10.8) μg/m3] than those in non-cases [PM2.5: 10.6 (SD 1.9) μg/m3; PM10: 15.7 (SD 2.5) μg/m3; NO2: 20.5 (SD 6.5) μg/m3; NOx: 31.3 (SD 12.3) μg/m3].

Table 1.

Characteristics of study participants.

Characteristics All participants Incident mental disorder cases Non-cases P
Number of participants 72,908 60,131 12,777
Sex [n (%)] <0.001
 Female 36,321 (49.8) 30,469 (50.7) 5852 (45.8)
 Male 36,587 (50.2) 29,662 (49.3) 6925 (54.2)
Age [year (mean ± SD)] 59.6 (7.2) 59.5 (7.1) 60.0 (7.4) <0.001
Ethnicity [n (%)] <0.001
 White 64,833 (88.9) 53,247 (88.6) 11,586 (90.7)
 Non-white 8075 (11.1) 6884 (11.4) 1191 (9.3)
Household income [n (%)] <0.001
 Less than 18,000 £ 18,222 (25.0) 14,056 (23.4) 4166 (32.6)
 18,000 to 30,999 £ 17,178 (23.6) 14,259 (23.7) 2919 (22.8)
 Greater than 31,000 £ 24,548 (33.7) 21,321 (35.5) 3227 (25.3)
 Unknown 12,960 (17.8) 10,495 (17.5) 2465 (19.3)
Alcohol intake [n (%)] <0.001
 Never 8782 (12.0) 7035 (11.7) 1747 (13.7)
 Occasional 20,439 (28.0) 16,706 (27.8) 3733 (29.2)
 Moderate 31,682 (43.5) 26,740 (44.5) 4942 (38.7)
 Heavy 12,005 (16.5) 9650 (16.0) 2355 (18.4)
Smoking status [n (%)] <0.001
 Never 37,270 (51.1) 32,928 (54.8) 4342 (34.0)
 Former 27,469 (37.7) 22,941 (38.2) 4528 (35.4)
 Current 8169 (11.2) 4262 (7.1) 3907 (30.6)
BMI [n (%)] <0.001
 Underweight 260 (0.4) 194 (0.3) 66 (0.5)
 Normal 13,940 (19.1) 11,716 (19.5) 2224 (17.4)
 Overweight 28,803 (39.5) 23,988 (39.9) 4815 (37.7)
 Obsess 29,905 (41.0) 24,233 (40.3) 5672 (44.4)
Residential area [n (%)] <0.001
 Urban 63,725 (87.4) 52,250 (86.9) 11,475 (89.8)
 Rural 9183 (12.6) 7881 (13.1) 1302 (10.2)
Physical activity [n (%)] <0.001
 Low 12,405 (17.0) 9992 (16.6) 2413 (18.9)
 Moderate 23,293 (31.9) 19,478 (32.4) 3815 (29.9)
 High 21,061 (28.9) 17,689 (29.4) 3372 (26.4)
 Unknown 16,149 (22.1) 12,972 (21.6) 3177 (24.9)
Educational level [n (%)] <0.001
 College or university degree 30,203 (41.4) 25,959 (43.2) 4244 (33.2)
 Any school degree 19,950 (27.4) 16,508 (27.5) 3442 (26.9)
 Vocational qualifications or other 22,755 (31.2) 17,664 (29.4) 5091 (39.8)
Time spent outdoors [n (%)] <0.001
 Low 28,697 (39.4) 24,215 (40.3) 4482 (35.1)
 Moderate 21,923 (30.1) 18,033 (30.0) 3890 (30.4)
 High 22,288 (30.6) 17,883 (29.7) 4405 (34.5)
Greenspace [n (%)] <0.001
 Low 24,060 (33.0) 19,673 (32.7) 4387 (34.3)
 Moderate 24,060 (33.0) 19,631 (32.6) 4429 (34.7)
 High 24,788 (34.0) 20,827 (34.6) 3961 (31.0)
Number of comorbidity diseases 0.7 (0.8) 0.7 (0.8) 0.8 (0.9) <0.001
Hypoglycemic medication [n (%)] <0.001
 No 56,396 (77.4) 47,068 (78.3) 9328 (73.0)
 Yes 16,512 (22.6) 13,063 (21.7) 3449 (27.0)
PM2.5 [μg/m3 (mean ± SD)] 9.9 (1.8) 9.8 (1.7) 10.6 (1.9) <0.001
PM10 [μg/m3 (mean ± SD)] 14.8 (2.5) 14.6 (2.4) 15.7 (2.5) <0.001
NO2 [μg/m3 (mean ± SD)] 18.3 (6.1) 17.8 (6.0) 20.5 (6.5) <0.001
NOx [μg/m3 (mean ± SD)] 27.4 (11.2) 26.5 (10.8) 31.3 (12.3) <0.001
Ambient temperature [°C (mean ± SD)] 9.8 (2.4) 9.8 (2.4) 9.9 (2.5) <0.001
Open in a new tab

Abbreviations: PM2.5, fine particulate matter; PM10, inhalable particulate matter; NO2, nitrogen dioxide; NOx, nitrogen oxides; SD, standard deviation.

3.2. Association estimated by non-weighted time-varying covariates Cox model

There were significant positive associations between the exposure to all four air pollutants and incident mental disorders among patients with prediabetes and diabetes estimated by the non-weighted time-varying covariates Cox model (Table 2). The estimates were more prominent in the diabetic participants than those with prediabetes. Among individuals with prediabetes, the HRs per IQR increase in PM2.5, PM10, NO2, and NOx were 1.16 (1.10, 1.22), 1.14 (1.09, 1.19), 1.15 (1.10, 1.19), and 1.12 (1.08, 1.16), respectively. Among individuals with diabetes, an IQR elevation in PM2.5, PM10, NO2, and NOx were associated with 1.20 (1.12, 1.28), 1.17 (1.10, 1.23), 1.16 (1.10, 1.22), and 1.15 (1.10, 1.20) fold increased risk of mental disorders. Similarly, significant positive associations were observed between all four air pollutants and incident anxiety disorders. In addition, air pollution was significantly associated with incident depressive disorders among patients with diabetes.

Table 2.

The associations between air pollution and incident mental disorders among people with prediabetes and diabetes estimated by the non-weighted time-varying covariates Cox model.

Pollutant Mental disorders
 Depressive disorders
 Anxiety disorders
HR (95 % CI) P HR (95 % CI) P HR (95 % CI) P
Total
PM2.5 1.18 (1.13, 1.22) <0.001 1.04 (0.96, 1.13) 0.332 1.31 (1.18, 1.46) <0.001
PM10 1.15 (1.11, 1.19) <0.001 1.03 (0.97, 1.11) 0.330 1.27 (1.18, 1.36) <0.001
NO2 1.15 (1.12, 1.19) <0.001 1.04 (0.98, 1.11) 0.179 1.18 (1.10, 1.27) <0.001
NOx 1.13 (1.10, 1.17) <0.001 1.03 (0.97, 1.09) 0.298 1.15 (1.08, 1.22) <0.001



Prediabetes
PM2.5 1.16 (1.10, 1.22) <0.001 1.06 (0.93, 1.21) 0.368 1.31 (1.18, 1.46) <0.001
PM10 1.14 (1.09, 1.19) <0.001 1.05 (0.94, 1.17) 0.377 1.26 (1.15, 1.38) <0.001
NO2 1.15 (1.10, 1.19) <0.001 1.03 (0.95, 1.12) 0.429 1.13 (1.01, 1.26) <0.001
NOx 1.12 (1.08, 1.16) <0.001 1.01 (0.94, 1.10) 0.720 1.11 (1.01, 1.23) 0.042



Diabetes
PM2.5 1.20 (1.12, 1.28) <0.001 1.14 (1.01, 1.29) 0.036 1.34 (1.17, 1.54) <0.001
PM10 1.17 (1.10, 1.23) <0.001 1.08 (1.01, 1.17) 0.045 1.27 (1.15, 1.38) <0.001
NO2 1.16 (1.10, 1.22) <0.001 1.08 (1.01, 1.17) 0.036 1.21 (1.11, 1.32) <0.001
NOx 1.15 (1.10, 1.20) <0.001 1.07 (1.01, 1.15) 0.046 1.17 (1.09, 1.27) <0.001
Open in a new tab

Abbreviations: PM2.5, fine particulate matter; PM10, inhalable particulate matter; NO2, nitrogen dioxide; NOx, nitrogen oxides; HR, hazard ratio.

3.3. Association estimated by IPW-weighted time-varying covariates Cox model

Fig. 1 shows the absolute correlation (AC) of covariates for unweighted and weighted populations. For unweighted populations, absolute correlations for the percentage of green space, ethnicity, Townsend deprivation index, and residential area were larger than 0.1. The averaged absolute correlations were summarized in Table S2. After weighting, the averaged absolute correlation values were reduced to below 0.1, which suggested a good covariate balance.

Fig. 1.

Fig. 1

Open in a new tab

Absolute correlation for unweighted and weighted populations (AC values of <0.1 indicate good covariate balance).

We found significant effects of exposure to air pollution on incident mental disorders estimated by the IPW weighted time-varying covariates Cox model. The exposure-response relationship curves for the association between air pollutants and the risk of incident mental disorders show a non-linear trend of flattening at low concentrations and relatively steepening at high concentrations (Fig. 2). Furthermore, the effects were higher in patients with diabetes than in prediabetes (Table 3). The HRs for incident mental disorders in patients with prediabetes were 1.18 (1.12, 1.24), 1.15 (1.10, 1.20), 1.18 (1.13, 1.23), and 1.15 (1.11, 1.19) per IQR increment in PM2.5, PM10, NO2, and NOx. In patients with diabetes, an IQR increase in PM2.5, PM10, NO2, and NOx was significantly associated with 1.21 (1.13, 1.29), 1.17 (1.11, 1.24), 1.19 (1.13, 1.25), and 1.17 (1.12, 1.23) fold increased risk of mental disorders. Similar results were found for anxiety disorders. However, only among patients with diabetes, the effect of air pollution on incident depressive disorders was significant.

Fig. 2.

Fig. 2

Open in a new tab

The exposure-response relationship curves for the association between air pollutants and the risk of incident mental disorders.

Table 3.

The associations between air pollution and incident mental disorders among people with prediabetes and diabetes estimated by the IPW weighted time-varying covariates Cox model.

Pollutant Mental disorders
 Depressive disorders
 Anxiety disorders
HR (95 % CI) P HR (95 % CI) P HR (95 % CI) P
Total
PM2.5 1.19 (1.14, 1.24) <0.001 1.14 (1.03, 1.25) 0.009 1.36 (1.24, 1.48) <0.001
PM10 1.16 (1.12, 1.20) <0.001 1.11 (1.02, 1.20) 0.014 1.29 (1.20, 1.39) <0.001
NO2 1.18 (1.15, 1.22) <0.001 1.12 (1.05, 1.20) 0.001 1.21 (1.13, 1.29) <0.001
NOx 1.16 (1.13, 1.19) <0.001 1.05 (0.98, 1.12) 0.151 1.18 (1.11, 1.25) <0.001



Prediabetes
PM2.5 1.18 (1.12, 1.24) <0.001 1.09 (0.95, 1.24) 0.204 1.33 (1.19, 1.49) <0.001
PM10 1.15 (1.10, 1.20) <0.001 1.07 (0.96, 1.19) 0.219 1.27 (1.16, 1.39) <0.001
NO2 1.18 (1.13, 1.23) <0.001 1.11 (1.00, 1.23) 0.061 1.19 (1.06, 1.33) <0.001
NOx 1.15 (1.11, 1.19) <0.001 1.04 (0.94, 1.15) 0.458 1.16 (1.05, 1.28) <0.001



Diabetes
PM2.5 1.21 (1.13, 1.29) <0.001 1.19 (1.04, 1.37) 0.014 1.38 (1.20, 1.59) <0.001
PM10 1.17 (1.11, 1.24) <0.001 1.14 (1.01, 1.28) 0.027 1.31 (1.17, 1.48) <0.001
NO2 1.19 (1.13, 1.25) <0.001 1.12 (1.02, 1.23) 0.016 1.21 (1.12, 1.32) <0.001
NOx 1.17 (1.12, 1.23) <0.001 1.09 (1.01, 1.19) 0.045 1.18 (1.10, 1.27) <0.001
Open in a new tab

Abbreviations: PM2.5, fine particulate matter; PM10, inhalable particulate matter; NO2, nitrogen dioxide; NOx, nitrogen oxides; HR, hazard ratio.

3.4. Stratified analyses

Fig. 3, Fig. 4 and Table S3 showed the effects of air pollution on incident mental disorders among people with prediabetes or diabetes stratified by sex, age, educational level, residential area, smoking status, and alcohol intake. The effects were stronger in the elderly aged >65 years than those aged ≤65 years (P for interaction: <0.001 for PM2.5, <0.001 for PM10, <0.001 for NO2, <0.001 for NOx). Likewise, the association between air pollution and incident mental disorders was more pronounced among people who were alcohol drinkers and living in urban areas (P for interaction for alcohol intake: 0.022 for PM2.5, 0.042 for NO2, 0.040 for NOx; P for interaction for residential area: 0.037 for NOx).

Fig. 3.

Fig. 3

Open in a new tab

Hazard ratios and 95 % confidence interval of incident mental disorders associated with per IQR increase in PM2.5 and PM10, stratified by sex, age, educational level, residential area, smoking status, and alcohol intake.

Fig. 4.

Fig. 4

Open in a new tab

Hazard ratios and 95 % confidence interval of incident mental disorders associated with per IQR increase in NO2 and NOx, stratified by sex, age, educational level, residential area, smoking status, and alcohol intake.

3.5. Sensitivity analyses

The results of the sensitivity analyses were summarized in Table S2 and Table S4. The E-value showed that the PM2.5 model was the most robust for unmeasured confounding. The results remained stable when we added the cubic spline for continuous variables, excluded participants with mental disorder events or death within 1 year of follow-up, excluded the participants with missing data on covariates, used the Cox proportional-hazards model and marginal structural model, and restricted the analyses in participants with air-pollution concentrations generally below the third quartile of the entire cohort throughout the study period. After the 1:1 propensity score matching, 22,589 pairs of participants with prediabetes and diabetes were included in the analyses. The results were similar to those of the main analyses.

4. Discussion

This study provided evidence that long-term exposure to PM2.5, PM10, NO2, and NOx had adverse effects on incident mental disorders among patients with prediabetes and diabetes. Furthermore, these effects were stronger in patients with diabetes than in prediabetes.

This was the first prospective cohort study to investigate the causal link between air pollution and incident mental disorders among people with prediabetes and diabetes. Although no previous studies have examined such associations using the causal modeling method, our findings can be supported by studies that used the traditional regression method. Numerous epidemiological studies have reported the association between air pollution and mental disorders in the general population (Gładka et al., 2018; Petrowski et al., 2021; Xue et al., 2019). For instance, recent study showed that long-term exposure to air pollution was associated with incident depression and anxiety (Yang et al., 2023). Furthermore, one time-series study illustrated that air pollutants including PM10, NO2, SO2, and CO significantly increased the risk of emergency department visits for depressive episodes in the population with diabetes mellitus (Cho et al., 2014).

Several mechanisms may explain the observed association between air pollution and mental disorders. One plausible pathway is that air pollutants could cross the blood-brain barrier, cause oxidative stress and inflammation in the central nervous system, and further lead to neurological and psychiatric disorders (Xu et al., 2016). This mechanism could be supported by findings from experimental research on animals. One study observed up-regulated hippocampal pro-inflammatory cytokine expression when mice inhaled PM2.5 (Fonken et al., 2011). Another study found that acute exposure to NO2 led to mitochondrial dysfunction in the rat cortex (Yan et al., 2015). Hyunyoung Kim et al. also reported that the effects of PM2.5 exposure on mitochondrial structure and function may result in cognitive decline and neurodegeneration (Kim et al., 2020).

We found significant effects of air pollution on anxiety disorders but not depression among the population with prediabetes, which is similar to a previous study (Motoc et al., 2023). The possible explanation is that some individual-level factors including biological factors may be more important in predicting depression. And previous studies indicated that depression may be more strongly related to loss, whereas anxiety may result from current and future threats (Eysenck et al., 2006). It should be noted that the non-statistically significant results found in the pre-diabetes population do not indicate that air pollution is not harmful to this population. The P-value is affected by many factors such as sample size, random error, and confounding factors. Moreover, the effect values in our study still suggest the harmful effects of air pollution. Therefore, this finding needs to be further verified and studied.

Moreover, diabetes and mental disorders can be partly ascribed to the disruption of the psycho-endocrine-immune system (Bui and de Vos, 2021; Figueira and Ouakinin, 2010). Previous studies suggested that air pollutants could activate the endocrine stress response by increasing glucocorticoid stress hormones and may further disturb the psycho-endocrine-immune system (Thomson, 2019). Therefore, patients with a disturbed psycho-endocrine-immune system may be more susceptible to other disorders owing to air pollution exposure (Cho et al., 2014). Nevertheless, the mechanisms of the effects of air pollution on mental disorders among people with prediabetes and diabetes need to be studied further.

Stratified analyses suggested that the effects of air pollution on incident mental disorders among people with prediabetes or diabetes were stronger in individuals who were old (>65 years), alcohol drinkers, or living in urban areas. Similar to the previous studies, the elderly was more sensitive to air pollution (Li et al., 2020; Lu et al., 2020). Because of the aging process, older bodies are less able to fend off inflammation, oxidative stress, and other damage from air pollution (Lim et al., 2012). At the same time, these damages are potential causes of mental disorders, which may contribute to the elderly's susceptibility (Tang et al., 2020). This finding suggests that protective measures, such as reducing time spent outdoors and wearing masks, should be taken for the elderly during the period of severe air pollution. Moreover, our findings are consistent with a previous study which indicated that health behavior such as drinking could modify the association between air pollution and mental health (Yang et al., 2021). We also found that the effects of air pollution were more pronounced in people living in urban areas. This may be because air pollution concentrations are higher in urban than in rural areas, and urban people live under more stress (Wheeler and Ben-Shlomo, 2005).

This study highlights the imperative for long-term adherence to public health policies aimed at reducing air pollution concentrations. It also suggests that policymakers should pay more attention to people with medical conditions, including prediabetes and diabetes. Specific measures ought to be implemented to protect these vulnerable populations from the adverse effects of air pollution. In addition, our findings showed that people with prediabetes also can be easily influenced by air pollution. However, previous research shows that the awareness rate of prediabetes is very low (Gopalan et al., 2015). As a result, this group of the population may neglect health management. Therefore, regular screening programs should be implemented to identify high-risk groups, and health education and intervention programs should be carried out to support individuals with prediabetes.

The current study has several strengths. First, it is based on a prospective cohort with a large sample size, which provides abundant information on demography, lifestyle, medical history, and environment. Second, we explored associations between air pollution and incident mental disorders by causal modeling approaches. The generalized propensity score method created a pseudo population which reduced the influence of measured confounders. Last, we studied multiple air pollutants and cause-specific mental disorders, which provided relatively comprehensive evidence for the association between air pollution and mental disorders among susceptible populations.

Several limitations should be acknowledged. The causal modeling method relies on the strong assumption that all confounders have been measured. While we cannot guarantee that no residual confounding remains, we did, however, include several confounders from the individual level and area level, and we also included calendar year as a surrogate for some unmeasured confounders which may change over time. We further calculated E-value to test the unmeasured confounders covaried with the exposure and outcome. In addition, patients without medical records may be ignored. Moreover, since the covariates except for air pollution, ambient temperature, and age were only reported at baseline, the potential for time-varying changes may influence subsequent mental disorder risk.

This study focused on the effects of air pollution on mental disorders among people with a medical condition in UK. It is unclear whether similar findings exist in populations of other countries or regions. And this study only focuses on the incidence of mental disorders. The effects of air pollution on disease progression and prognosis, as well as the health and economic benefits of policies to reduce air pollution on mental health, remain to be investigated. Furthermore, future researchers could advance the causal modeling method to simulate the randomized controlled trial for more reliable effect estimates.

5. Conclusions

In conclusion, we find potential causal link between air pollution and incident mental disorders among patients with prediabetes and diabetes. Reducing air pollution exposure would have great public health benefits for these vulnerable people by decreasing the number of mental-disorder cases. These findings suggest the importance of adherence to public health policies to control air pollution.

Funding

This work was supported by the Bill & Melinda Gates Foundation [Grant Number: INV-016826].

Role of the funding source

The funders had no role in study design, data collation, data analysis, data interpretation, writing of the manuscript, and decision to submit. The corresponding author had full access to all of the data and the final responsibility for the decision to submit for publication.

CRediT authorship contribution statement

J.F. and M.C. conceived and designed the study. S.Z., Y.Y., and G.C. performed data analysis. J.H. was responsible for visualization. X.W. provided statistical expertise. Z.Q., S.E.M., C.W. and M.T. revised the manuscript. H.L. supervised the study. All authors contributed to the interpretation of the results and critical revision of the manuscript. All authors read and approved the final manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors are thankful to the staff and participants of the UK Biobank. This research was conducted using the UK Biobank resource under application number 69550.

Editor: Lidia Minguez Alarcon

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.scitotenv.2023.165235.

Appendix A. Supplementary data

Supplementary materials

mmc1.docx (299.1KB, docx)

Data availability

The authors do not have permission to share data.

References

  1. Andersen P.K., Gill R.D. Cox’s regression model for counting processes: a large sample study. Ann. Stat. 1982;10(1100–1120):21. [Google Scholar]
  2. Arias D., Saxena S., Verguet S. Quantifying the global burden of mental disorders and their economic value. EClinicalMedicine. 2022;54 doi: 10.1016/j.eclinm.2022.101675. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Atkins J.L., Masoli J.A.H., Delgado J., Pilling L.C., Kuo C.L., Kuchel G.A., et al. Preexisting comorbidities predicting COVID-19 and mortality in the UK biobank community cohort. J. Gerontol. A Biol. Sci. Med. Sci. 2020;75:2224–2230. doi: 10.1093/gerona/glaa183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Austin P.C. Assessing covariate balance when using the generalized propensity score with quantitative or continuous exposures. Stat. Methods Med. Res. 2019;28:1365–1377. doi: 10.1177/0962280218756159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Baumeister H., Hutter N., Bengel J., Härter M. Quality of life in medically ill persons with comorbid mental disorders: a systematic review and meta-analysis. Psychother. Psychosom. 2011;80:275–286. doi: 10.1159/000323404. [DOI] [PubMed] [Google Scholar]
  6. Benton M., Cleal B., Prina M., Baykoca J., Willaing I., Price H., et al. Prevalence of mental disorders in people living with type 1 diabetes: a systematic literature review and meta-analysis. Gen. Hosp. Psychiatry. 2023;80:1–16. doi: 10.1016/j.genhosppsych.2022.11.004. [DOI] [PubMed] [Google Scholar]
  7. Boden M.T. Prevalence of mental disorders and related functioning and treatment engagement among people with diabetes. J. Psychosom. Res. 2018;106:62–69. doi: 10.1016/j.jpsychores.2018.01.001. [DOI] [PubMed] [Google Scholar]
  8. Buchberger B., Huppertz H., Krabbe L., Lux B., Mattivi J.T., Siafarikas A. Symptoms of depression and anxiety in youth with type 1 diabetes: a systematic review and meta-analysis. Psychoneuroendocrinology. 2016;70:70–84. doi: 10.1016/j.psyneuen.2016.04.019. [DOI] [PubMed] [Google Scholar]
  9. Bui T.P.N., de Vos W.M. Next-generation therapeutic bacteria for treatment of obesity, diabetes, and other endocrine diseases. Best Pract. Res. Clin. Endocrinol. Metab. 2021;35 doi: 10.1016/j.beem.2021.101504. [DOI] [PubMed] [Google Scholar]
  10. Cai M., Xie Y., Bowe B., Gibson A.K., Zayed M.A., Li T., et al. Temporal trends in incidence rates of lower extremity amputation and associated risk factors among patients using veterans health administration services from 2008 to 2018. JAMA Netw. Open. 2021;4 doi: 10.1001/jamanetworkopen.2020.33953. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cai M., Li H., Wu Y., Zhang S., Wang X., Zhang Z., et al. Ambient air pollution associated with body fat percentages at different body compartments: a cohort study of UK biobank participants. Environ. Health Perspect. 2022;130:67702. doi: 10.1289/EHP10920. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Cai M., Lin X., Wang X., Zhang S., Wang C., Zhang Z., et al. Long-term exposure to ambient fine particulate matter chemical composition and in-hospital case fatality among patients with stroke in China. Lancet Reg. Health West Pac. 2023;32 doi: 10.1016/j.lanwpc.2022.100679. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Caleyachetty R., Littlejohns T., Lacey B., Bešević J., Conroy M., Collins R., et al. United Kingdom biobank (UK biobank): JACC focus seminar 6/8. J. Am. Coll. Cardiol. 2021;78:56–65. doi: 10.1016/j.jacc.2021.03.342. [DOI] [PubMed] [Google Scholar]
  14. Cho J., Choi Y.J., Suh M., Sohn J., Kim H., Cho S.K., et al. Air pollution as a risk factor for depressive episode in patients with cardiovascular disease, diabetes mellitus, or asthma. J. Affect. Disord. 2014;157:45–51. doi: 10.1016/j.jad.2014.01.002. [DOI] [PubMed] [Google Scholar]
  15. Cobham V.E., Hickling A., Kimball H., Thomas H.J., Scott J.G., Middeldorp C.M. Systematic review: anxiety in children and adolescents with chronic medical conditions. J. Am. Acad. Child Adolesc. Psychiatry. 2020;59:595–618. doi: 10.1016/j.jaac.2019.10.010. [DOI] [PubMed] [Google Scholar]
  16. Collaborators G.M.D. Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Psychiatry. 2022;9:137–150. doi: 10.1016/S2215-0366(21)00395-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dan Hollis M.M., Kendon Michael, Legg Tim, Simpson Ian. HadUK-grid—A new UK dataset of gridded climate observations. Geosci. Data J. 2019;6:151–159. [Google Scholar]
  18. Egede L.E., Nietert P.J., Zheng D. Depression and all-cause and coronary heart disease mortality among adults with and without diabetes. Diabetes Care. 2005;28:1339–1345. doi: 10.2337/diacare.28.6.1339. [DOI] [PubMed] [Google Scholar]
  19. Eysenck M., Payne S., Santos R. Anxiety and depression: past, present, and future events. Cognit. Emot. 2006;20:274–294. [Google Scholar]
  20. Figueira M.L., Ouakinin S. Gender-related endocrinological dysfunction and mental disorders. Curr. Opin. Psychiatry. 2010;23:369–372. doi: 10.1097/YCO.0b013e3283399b86. [DOI] [PubMed] [Google Scholar]
  21. Fonken L.K., Xu X., Weil Z.M., Chen G., Sun Q., Rajagopalan S., et al. Air pollution impairs cognition, provokes depressive-like behaviors and alters hippocampal cytokine expression and morphology. Mol. Psychiatry. 2011;16(987–95):973. doi: 10.1038/mp.2011.76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Fu Y., Gu M., Wang R., Xu J., Sun S., Zhang H., et al. Abnormal functional connectivity of the frontostriatal circuits in type 2 diabetes mellitus. Front. Aging Neurosci. 2022;14:1055172. doi: 10.3389/fnagi.2022.1055172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Gao Q., Zhang Y., Liang J., Sun H., Wang T. High-dimensional generalized propensity score with application to omics data. Brief. Bioinform. 2021;22 doi: 10.1093/bib/bbab331. [DOI] [PubMed] [Google Scholar]
  24. Garrett C., Doherty A. Diabetes and mental health. Clin. Med. (Lond.) 2014;14:669–672. doi: 10.7861/clinmedicine.14-6-669. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Gładka A., Rymaszewska J., Zatoński T. Impact of air pollution on depression and suicide. Int. J. Occup. Med. Environ. Health. 2018;31:711–721. doi: 10.13075/ijomeh.1896.01277. [DOI] [PubMed] [Google Scholar]
  26. Gopalan A., Lorincz I.S., Wirtalla C., Marcus S.C., Long J.A. Awareness of prediabetes and engagement in diabetes risk–reducing behaviors. Am. J. Prev. Med. 2015;49:512–519. doi: 10.1016/j.amepre.2015.03.007. [DOI] [PubMed] [Google Scholar]
  27. Hao G., Zuo L., Xiong P., Chen L., Liang X., Jing C. Associations of PM2.5 and road traffic noise with mental health: Evidence from UK Biobank. Environ. Res. 2022;207 doi: 10.1016/j.envres.2021.112221. [DOI] [PubMed] [Google Scholar]
  28. Heine-Bröring R.C., Winkels R.M., Botma A., Wahab P.J., Tan A.C., Nagengast F.M., et al. Dietary supplement use is not associated with recurrence of colorectal adenomas: a prospective cohort study. Int. J. Cancer. 2013;132:666–675. doi: 10.1002/ijc.27647. [DOI] [PubMed] [Google Scholar]
  29. Hirano Keisuke, Imbens G.W. The propensity score with continuous treatments. Applied Bayesian Modeling and Causal Inference from Incomplete-Data. Perspectives. 2004;226164:73–84. [Google Scholar]
  30. Imbens G.W. The role of the propensity score in estimating dose-response functions. BIOMETRIKA. 2000;87:706–710. [Google Scholar]
  31. International Diabetes Federation. IDF Diabetes Atlas IDF, 10th edition, 2021. Available from https://www.diabetesatlas.org.
  32. Ioannidis J.P.A., Tan Y.J., Blum M.R. Limitations and misinterpretations of E-values for sensitivity analyses of observational studies. Ann. Intern. Med. 2019;170:108–111. doi: 10.7326/M18-2159. [DOI] [PubMed] [Google Scholar]
  33. Kim H., Kim W.H., Kim Y.Y., Park H.Y. Air pollution and central nervous system disease: a review of the impact of fine particulate matter on neurological disorders. Front. Public Health. 2020;8 doi: 10.3389/fpubh.2020.575330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kotwas A., Karakiewicz B., Zabielska P., Wieder-Huszla S., Jurczak A. Epidemiological factors for type 2 diabetes mellitus: evidence from the Global Burden of Disease. Arch. Public Health. 2021;79:110. doi: 10.1186/s13690-021-00632-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Li H., Zhang S., Qian Z.M., Xie X.H., Luo Y., Han R., et al. Short-term effects of air pollution on cause-specific mental disorders in three subtropical Chinese cities. Environ. Res. 2020;191 doi: 10.1016/j.envres.2020.110214. [DOI] [PubMed] [Google Scholar]
  36. Lim Y.H., Kim H., Kim J.H., Bae S., Park H.Y., Hong Y.C. Air pollution and symptoms of depression in elderly adults. Environ. Health Perspect. 2012;120:1023–1028. doi: 10.1289/ehp.1104100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Lu P., Zhang Y., Xia G., Zhang W., Xu R., Wang C., et al. Attributable risks associated with hospital outpatient visits for mental disorders due to air pollution: a multi-city study in China. Environ. Int. 2020;143 doi: 10.1016/j.envint.2020.105906. [DOI] [PubMed] [Google Scholar]
  38. Majidi S., Driscoll K.A., Raymond J.K. Anxiety in children and adolescents with type 1 diabetes. Curr. Diab. Rep. 2015;15:47. doi: 10.1007/s11892-015-0619-0. [DOI] [PubMed] [Google Scholar]
  39. Motoc I., Hoogendijk E.O., Timmermans E.J., Deeg D., Penninx B., Huisman M. Social and physical neighbourhood characteristics and 10-year incidence of depression and anxiety in older adults: results from the Longitudinal Aging Study Amsterdam. Soc. Sci. Med. 2023;327 doi: 10.1016/j.socscimed.2023.115963. [DOI] [PubMed] [Google Scholar]
  40. Nissen A., Hynek K.A., Scales D., Hilden P.K., Straiton M. Chronic pain, mental health and functional impairment in adult refugees from Syria resettled in Norway: a cross-sectional study. BMC Psychiatry. 2022;22:571. doi: 10.1186/s12888-022-04200-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Pan C., Ye J., Wen Y., Chu X., Jia Y., Cheng B., et al. The associations between sleep behaviors, lifestyle factors, genetic risk and mental disorders: a cohort study of 402 290 UK Biobank participants. Psychiatry Res. 2022;311 doi: 10.1016/j.psychres.2022.114488. [DOI] [PubMed] [Google Scholar]
  42. Perry M., Hollis D. The development of a new set of long-term climate averages for the UK. Int. J. Climatol. 2005;25:1023–1039. [Google Scholar]
  43. Petrowski K., Bührer S., Strauß B., Decker O., Brähler E. Examining air pollution (PM(10)), mental health and well-being in a representative German sample. Sci. Rep. 2021;11:18436. doi: 10.1038/s41598-021-93773-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Roberts S., Arseneault L., Barratt B., Beevers S., Danese A., Odgers C.L., et al. Exploration of NO(2) and PM(2.5) air pollution and mental health problems using high-resolution data in London-based children from a UK longitudinal cohort study. Psychiatry Res. 2019;272:8–17. doi: 10.1016/j.psychres.2018.12.050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Roy T., Lloyd C.E. Epidemiology of depression and diabetes: a systematic review. J. Affect. Disord. 2012;142:S8–S21. doi: 10.1016/S0165-0327(12)70004-6. [DOI] [PubMed] [Google Scholar]
  46. Safiri S., Karamzad N., Kaufman J.S., Bell A.W., Nejadghaderi S.A., Sullman M.J.M., et al. Prevalence, deaths and disability-adjusted-life-years (DALYs) due to type 2 diabetes and its attributable risk factors in 204 countries and territories, 1990-2019: results from the global burden of disease study 2019. Front. Endocrinol. (Lausanne) 2022;13 doi: 10.3389/fendo.2022.838027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Tang M.L., Li D., Liew Z., Wei F., Wang J.B., Jin M.J., et al. The association of short-term effects of air pollution and sleep disorders among elderly residents in China. Sci. Total Environ. 2020:708. doi: 10.1016/j.scitotenv.2019.134846. [DOI] [PubMed] [Google Scholar]
  48. Thomson E.M. Air pollution, stress, and allostatic load: linking systemic and central nervous system impacts. J. Alzheimers Dis. 2019;69:597–614. doi: 10.3233/JAD-190015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. UK Air. Modelled Background Pollution Data. London: Department for Environment, Food and Rural Affairs. https://uk-air.defra.gov.uk/data/pcm-data., 2022.
  50. VanderWeele T.J., Ding P. Sensitivity analysis in observational research: introducing the E-value. Ann. Intern. Med. 2017;167:268–274. doi: 10.7326/M16-2607. [DOI] [PubMed] [Google Scholar]
  51. Ventriglio A., Bellomo A., di Gioia I., Di Sabatino D., Favale D., De Berardis D., et al. Environmental pollution and mental health: a narrative review of literature. CNS Spectr. 2021;26:51–61. doi: 10.1017/S1092852920001303. [DOI] [PubMed] [Google Scholar]
  52. Wang Y., Lee M., Liu P., Shi L., Yu Z., Abu Awad Y., et al. Doubly robust additive hazards models to estimate effects of a continuous exposure on survival. Epidemiology. 2017;28:771–779. doi: 10.1097/EDE.0000000000000742. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Wang Y., Wei J., Zhang Y., Guo T., Chen S., Wu W., et al. Estimating causal links of long-term exposure to particulate matters with all-cause mortality in South China. Environ. Int. 2023;171 doi: 10.1016/j.envint.2022.107726. [DOI] [PubMed] [Google Scholar]
  54. Wheeler B.W., Ben-Shlomo Y. Environmental equity, air quality, socioeconomic status, and respiratory health: a linkage analysis of routine data from the Health Survey for England. J. Epidemiol. Community Health. 2005;59:948–954. doi: 10.1136/jech.2005.036418. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Wu X., Braun D., Schwartz J., Kioumourtzoglou M.A., Dominici F. Evaluating the impact of long-term exposure to fine particulate matter on mortality among the elderly. Sci. Adv. 2020;6:eaba5692. doi: 10.1126/sciadv.aba5692. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Wu Y., Zhang S., Qian S.E., Cai M., Li H., Wang C., et al. Ambient air pollution associated with incidence and dynamic progression of type 2 diabetes: a trajectory analysis of a population-based cohort. BMC Med. 2022;20:375. doi: 10.1186/s12916-022-02573-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Xu S., Shetterly S., Raebel M.A., Ho P.M., Tsai T.T., Magid D. Estimating the effects of time-varying exposures in observational studies using Cox models with stabilized weights adjustment. Pharmacoepidemiol. Drug Saf. 2014;23:812–818. doi: 10.1002/pds.3601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Xu X., Ha S.U., Basnet R. A review of epidemiological research on adverse neurological effects of exposure to ambient air pollution. Front. Public Health. 2016;4:157. doi: 10.3389/fpubh.2016.00157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Xue T., Zhu T., Zheng Y., Zhang Q. Declines in mental health associated with air pollution and temperature variability in China. Nat. Commun. 2019;10:2165. doi: 10.1038/s41467-019-10196-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Yan W., Ji X., Shi J., Li G., Sang N. Acute nitrogen dioxide inhalation induces mitochondrial dysfunction in rat brain. Environ. Res. 2015;138:416–424. doi: 10.1016/j.envres.2015.02.022. [DOI] [PubMed] [Google Scholar]
  61. Yang T., Wang J., Huang J., Kelly F.J., Li G. Long-term exposure to multiple ambient air pollutants and association with incident depression and anxiety. JAMA Psychiatry. 2023;80:305–313. doi: 10.1001/jamapsychiatry.2022.4812. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Yang Y., Li R., Cai M., Wang X., Li H., Wu Y., et al. Ambient air pollution, bone mineral density and osteoporosis: results from a national population-based cohort study. Chemosphere. 2022;310 doi: 10.1016/j.chemosphere.2022.136871. [DOI] [PubMed] [Google Scholar]
  63. Yang Z.M., Song Q.H., Li J., Zhang Y.Q., Yuan X.C., Wang W.Q., et al. Air pollution and mental health: the moderator effect of health behaviors. Environ. Res. Lett. 2021;16 [Google Scholar]
  64. Zhang P., Guo D., Xu B., Huang C., Yang S., Wang W., et al. Association of serum 25-hydroxyvitamin D with cardiovascular outcomes and all-cause mortality in individuals with prediabetes and diabetes: results from the UK biobank prospective cohort study. Diabetes Care. 2022;45:1219–1229. doi: 10.2337/dc21-2193. [DOI] [PubMed] [Google Scholar]
  65. Zhang W., Gao C., Qing Z., Zhang Z., Bi Y., Zeng W., et al. Hippocampal subfields atrophy contribute more to cognitive impairment in middle-aged patients with type 2 diabetes rather than microvascular lesions. Acta Diabetol. 2021;58:1023–1033. doi: 10.1007/s00592-020-01670-x. [DOI] [PubMed] [Google Scholar]
  66. Zhang Z., Chen L., Qian Z.M., Li H., Cai M., Wang X., et al. Residential green and blue space associated with lower risk of adult-onset inflammatory bowel disease: findings from a large prospective cohort study. Environ. Int. 2022;160 doi: 10.1016/j.envint.2022.107084. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary materials

mmc1.docx (299.1KB, docx)

Data Availability Statement

The authors do not have permission to share data.

RESOURCES