Ambient air pollution exposure accelerates the occurrence of 78 non-communicable chronic diseases: an accelerated failure time analysis of a nationwide cohort

Abstract

Ambient air pollution is a well-established risk factor for chronic diseases, but its impact on disease onset age remains unclear. This study systematically evaluated the acceleration effect of air pollutants on the onset of 78 chronic diseases using over 900,000 hospitalization records from 396,000 UK Biobank participants. Both particulate matter and nitrogen oxides were associated with accelerated onset of 46 out of 78 diseases (9 cardiovascular diseases, 7 respiratory diseases, 14 psychological/neurological disorders, 3 digestive diseases, 2 cancers, and 11 other chronic diseases). Significant associations including those for common chronic diseases were observed. Each interquartile range (IQR) increase in PM_2.5 was strongly associated with a 0.93% (95% CI—0.86%, 1.00%) decrease in age at onset (AAO) of hypertension. Similarly, NO_x was associated with a 0.96% (95% CI—0.82%, 1.09%) decrease in AAO of COPD, PM₁₀ with a 0.95% (95% CI—0.81%, 1.09%) decrease in AAO of diabetes, and NO₂ with a 0.88% (95% CI—0.77%, 1.00%) decrease in AAO of dementia. Notably, we observed that neurological/psychological disorders were observed to be mostly affected, including schizophrenia, dystonia, polyneuropathies, and migraine, with 1 ~ 3% reduction in the AAO. On a population level, PM_2.5 overexposure (exceeding the WHO guideline of 5 μg/m³) accounted for 539,320 person-years of accelerated AAO across 78 chronic conditions, with hypertension (18.10%), asthma (6.03%), and diabetes (5.39%) contributing the most. This study provides the first evidence that air pollutants could accelerate onset of common chronic diseases. Findings highlight the urgent need for measures to improve air quality to slow progression of disease development.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s11357-025-01806-3.

Introduction

Non-communicable diseases (NCDs) are the leading cause of global morbidity and mortality, accounting for 74% of deaths worldwide in 2023 [1]. These diseases impose a significant burden on healthcare systems and economies, driving up healthcare costs and productivity, with the ongoing global aging trend further amplifying their impact [2, 3]. Chronic diseases are characterized by prolonged duration and slow progression. While these chronic conditions, such as cardiovascular disease (CVD), have historically been more prevalent among middle-aged and elderly populations, mounting evidence supports an alarming rise in their occurrence among younger individuals [4–6]. This trend is particularly evident in CVD (including hypertension and coronary artery disease) [7], colorectal cancer [8], and type 2 diabetes (T2D) [9]. The decreasing age of onset has emerged as a significant public health concern.

Studies have documented strong associations between ambient air pollution and chronic diseases [10–15], underscoring the pervasive impact of air pollution as a major environmental risk factor. However, there is still a limited understanding of whether, and to what extent, air pollutants may hasten the onset of chronic diseases. Exposure to air pollutants has been shown to trigger oxidative stress, systemic inflammation, and other pathophysiological changes that may accelerate the development of these conditions [16–18]. Gaining insights into how air pollution affects the timing of disease onset is crucial for informing effective prevention and intervention strategies.

This study aimed to investigate the potential effects of air pollution on the onset age of a broad spectrum of chronic diseases. Using data from 396,265 UK Biobank participants, the analysis examined 78 chronic diseases. Findings from this analysis would provide important evidence on the environmental risk factor responsible for the tendency toward younger occurrence of major chronic diseases.

Methods

Study population

This analysis used data from the UK Biobank, a population-based prospective cohort of more than 500,000 individuals who were 39–70 years old at the start of the study. Details about the study protocol and design has been published previously (18). All participants provided informed consent at recruitment. The North West Multicenter Research Ethics Committee (REC reference: 16/NW/0274) approved the UK Biobank.

Air pollution assessment

Concentrations of air pollutants were obtained from the UK AIR platform (https://uk-air.defra.gov.uk/data/pcm-data) of the Department for Environment, Food and Rural Affairs (DEFRA). This high-resolution near-surface data (1 km × 1 km) was provided through an air dispersion model and has been widely used in health effect analyses [19–21]. This model was created using data from the National Atmospheric Emissions Inventory (NAEI), inorganic aerosols, and dust resuspension. Background data from DEFRA’s Automatic Urban and Rural Network measurements were used to calibrate the model. DEFRA previously determined model reliability, with strong agreement between measured data and model estimates [19–21]. Model performance statistics can be found on DEFRA’s website. The ambient air pollution concentrations of PM_2.5, PM₁₀, NO₂, and NO_x for participants were estimated by linking their baseline residential coordinates. Each individual’s exposure window was defined by calculating the average air pollutant levels over the 4 years prior to the baseline survey, and the exposure levels were obtained by weighting the time spent at each residential address according to the duration of residence.

Ascertainment of chronic diseases

Participant endpoints were determined as date of initial chronic disease diagnosis, date of loss-to-follow-up, death, or final study follow-up (September 30, 2021 in England; July 31, 2021 in Scotland; February 28, 2018 in Wales), whichever first occurred.

The UK Biobank gathered chronic disease diagnoses and medical data from self-reports, hospital admission records, and death registration records provided by the UK National Health Services (NHS). For this analysis, 78 chronic diseases were selected, encompassing cardiovascular diseases (CVD), respiratory diseases (RD), digestive disorders, neurological and psychological disorders, cancer, and other chronic conditions. International Classification of Diseases (ICD)−9 and ICD-10 codes were used to identify primary and secondary diagnoses for participants admitted to hospital (Supplementary Table 1). Participants who had a diagnosis of any disease of interest at baseline were excluded from the study.

Assessment of covariates

Sociodemographic and behavioral covariates of interest included age, sex, ethnicity, highest level of education, body mass index (BMI), physical activity, alcohol consumption, smoking status, and Townsend deprivation index (TDI). Educational levels were classified into college education, any school degree lower than college, vocational qualifications, and other. Smoking status was categorized into never, former, and current. Alcohol consumption was classified as never, occasional, moderate, or heavy drinkers based on daily intake. Physical activity levels were determined using the self-reported International Physical Activity Questionnaire (IPAQ), which uses walking, moderate-intensity, and vigorous-intensity activities to classify overall activity levels as low (< 600 MET-min/week), moderate (600–3000 MET-min/week), and high (≥ 3000 MET-min/week). TDI was used to assess socioeconomic status based on a combination of social class, car ownership, household overcrowding, home ownership, and employment status, with higher score signaling a greater level of deprivation.

Statistical analysis

Characteristics of participants at baseline were described using mean ± (SDs) or frequency (percentage) by quintiles of air pollution and differences examined using ANOVA or χ² as appropriate.

We initially fitted Cox proportional hazards regression models to investigate the associations between air pollution exposure and incident risk of various chronic diseases. In the models, we adjusted for age, sex, ethnicity education level, body mass index, smoking status, alcohol consumption, physical activity, and TDI (included as quantiles). To account for multiple comparisons, statistical significance was determined using the Benjamini–Hochberg false discovery rate (FDR) correction method.

Accelerated failure time (AFT) regression models were run to assess the association between air pollution and age at onset of chronic diseases. Multivariable AFT models adjusted for the same covariates as Cox regression models. AFT models are a class of parametric survival analysis models that directly relate survival time of an event to an exposure of interest [22, 23]. The AFT model was selected over the Cox proportional hazards model due to its direct estimation of time ratios (TRs), which quantify how air pollution accelerates or delays disease onset [24]. This approach aligns closely with our research aim of evaluating temporal shifts in disease occurrence. Compared to the semi-parametric Cox model, the parametric AFT framework can offer more interpretable and efficient estimates, not requiring proportional hazards [24, 25]. After examining Gaussian, exponential, Weibull, log-logistic, and log-normal distributions, Akaike information criterion (AIC) was selected to determine the best distribution (Supplemental Table 2). Weibull distribution was selected due to the best model fit. The exponentiated AFT regression coefficient represents acceleration factors (AF). Acceleration factor or time ratio (TR) and 95% confidence intervals (CI) of air pollutant were calculated for each interquartile range (IQR) increase. A time ratio of 1.0 indicates no association between air pollution exposure variable and chronic disease diagnosis. Values less than 1.0 indicate that air pollution decreases the time to chronic disease diagnosis (i.e., the event occurs more quickly), whereas values greater than 1.0 suggest that air pollution exposure increases the time chronic disease diagnosis. In addition, the dose–response relationship of air pollution and age at onset of diseases was fitted using a natural spline with three degrees. Age at onset of chronic diseases was predicted for high vs low levels of air pollution, with the median serving as the threshold. Furthermore, accelerated age of disease onset attributable to air pollution overexposure was calculated using a well-established method of attributable risk assessment. The formula for accelerated AAO is given by:

$$AcceleratedAAO=\frac{1}{n}{\sum }_{i=1}^n\frac{\mathit{AAO}}{{\mathit{TR}}^{(A_i-A_0)/IQR}}-AAO$$

[26], in which $i$ represents participants diagnosed with chronic diseases, $\mathit{TR}$ denotes the acceleration factor or time ratio related to the impact of air pollution on the age of onset for these diseases, $A_i$ is the annual average air pollution concentration for participants with chronic diseases, and $A_0$ is the recommended annual air pollution concentration based on the 2021 WHO Global Air Quality Guidelines. According to WHO air guideline, the recommended limits are 5 μg/m³ for PM_2.5, 15 μg/m³ for PM₁₀, and 10 μg/m³ for NO₂. Since no specific guideline exists for NO_x, an annual average of 25 μg/m³ was used as a reference. In this context, the accelerated AAO reflects the advancement in the age of onset of chronic diseases due to air pollution levels that exceed WHO’s recommended exposure limits. Finally, population attributable fractions (PAFs) were calculated for each chronic disease due to a high air pollution level using a well-established approach developed by Eide and Gefeller [27] and the “averisk” package in R software [28].

Stratified analyses and sensitivity analyses

Stratified analyses for associations between air pollutants and onset age of chronic diseases were conducted for participant age (< 60 years and ≥ 60 years) and sex (male and female). Sensitivity analyses examined the robustness of the results. First, participants diagnosed with a corresponding disease within the first 3 years of follow-up were excluded due to impact of reverse causality. Second, the main analyses were repeated with altered exposure windows over the 1 year, 2 years, and 3 years prior to baseline. Third, Fine and Grey competing risk models were generated to account for the influence of competing mortality risks.

Results

Descriptive results

A total of 396,265 participants with complete data were eligible for analysis. The mean age was 56.3 years, and 47.6% were male. The median (IQR) of annual average concentrations of PM_2.5, PM₁₀, NO₂, and NO_x were 10.7 (2.8) µg/m³, 16.9 (3.8) µg/m³, 20.3 (7.8) µg/m³, and 29.7 (15.5) µg/m³, respectively. Baseline characteristics of the participants stratified by air pollution levels (in quintile) are presented in Table 1. A greater portion of participants with higher pollution exposure were younger, non-White, highly deprived, smokers, and less physical active.

Table 1.

Baseline characteristics across quintiles of air pollution

	PM2.5		PM10		NO2		NOx
	Quintile 1	Quintile 5	Quintile 1	Quintile 5	Quintile 1	Quintile 5	Quintile 1	Quintile 5
Age, years	57.02 (7.94)	55.73 (8.28)	57.12 (7.91)	55.63 (8.30)	57.31 (7.84)	55.56 (8.33)	57.28 (7.84)	55.56 (8.33)
Sex, female, n (%)	55,371 (55.10)	54,902 (54.64)	55,321 (55.06)	54,562 (54.30)	55,206 (54.94)	54,203 (53.95)	55,228 (54.96)	54,129 (53.87)
Ethnicity, White, n (%)	98,930 (98.76)	83,167 (83.79)	98,985 (98.80)	83,302 (83.90)	99,044 (98.88)	83,307 (83.87)	99,061 (98.88)	83,351 (83.91)
BMI level, n (%)
Underweight	31,901 (31.87)	35,287 (35.50)	32,293 (32.26)	34,624 (34.84)	33,720 (33.68)	33,307 (33.53)	33,738 (33.70)	33,206 (33.44)
Normal	471 (0.47)	723 (0.73)	457 (0.46)	712 (0.72)	471 (0.47)	697 (0.70)	462 (0.46)	702 (0.71)
Overweight	43,585 (43.55)	39,823 (40.06)	43,615 (43.57)	39,993 (40.24)	43,387 (43.33)	40,132 (40.41)	43,374 (43.32)	40,115 (40.39)
Obese	24,127 (24.11)	23,565 (23.71)	23,731 (23.71)	24,062 (24.21)	22,550 (22.52)	25,185 (25.36)	22,552 (22.52)	25,292 (25.47)
Townsend deprivation index	− 1.82 (2.87)	0.57 (3.30)	− 2.00 (2.78)	0.61 (3.31)	− 2.59 (2.28)	0.99 (3.32)	− 2.61 (2.27)	0.99 (3.33)
Higher education, n (%)	47,984 (48.21)	51,595 (54.00)	48,590 (48.81)	51,540 (52.81)	51,357 (51.57)	47,865 (50.15)	51,162 (51.36)	47,524 (49.81)
Intake of alcohol, n (%)
Never	7049 (7.03)	11,152 (11.18)	6803 (6.78)	11,300 (11.33)	6090 (6.07)	11,740 (11.77)	6129 (6.11)	11,793 (11.83)
Occasional	21,610 (21.54)	24,274 (24.34)	21,419 (21.35)	24,324 (24.39)	20,297 (20.22)	24,634 (24.70)	20,408 (20.33)	24,682 (24.75)
Moderate	52,828 (52.66)	42,036 (42.15)	52,852 (52.67)	42,258 (42.37)	51,390 (51.20)	42,750 (42.86)	51,286 (51.10)	42,709 (42.83)
Heavy	18,837 (18.78)	22,268 (22.33)	19,263 (19.20)	21,851 (21.91)	22,587 (22.51)	20,608 (20.66)	22,546 (22.46)	20,543 (20.60)
Smoking status, n (%)
Never	56,871 (56.64)	52,614 (52.62)	57,232 (56.99)	52,370 (52.37)	57,494 (57.25)	51,354 (51.35)	57,586 (57.34)	51,307 (51.30)
Previous	33,743 (33.60)	34,204 (34.21)	33,727 (33.59)	34,229 (34.23)	34,679 (34.53)	34,171 (34.17)	34,552 (34.40)	34,165 (34.16)
Current	9478 (9.44)	12,615 (12.62)	9145 (9.11)	12,828 (12.83)	7967 (7.93)	13,888 (13.89)	8005 (7.97)	13,934 (13.93)
Prefer not to answer	321 (0.32)	554 (0.55)	315 (0.31)	569 (0.57)	284 (0.28)	591 (0.59)	286 (0.28)	599 (0.60)
Physical activity level, n (%)
Low	14,441 (17.74)	14,433 (18.17)	14,408 (17.67)	14,786 (18.32)	14,701 (17.82)	14,757 (18.86)	14,747 (17.88)	14,750 (18.86)
Moderate	33,435 (41.07)	33,887 (42.67)	33,300 (40.83)	34,159 (42.32)	33,399 (40.48)	32,652 (41.73)	33,369 (40.47)	32,574 (41.66)
High	33,536 (41.19)	31,098 (39.16)	33,841 (41.50)	31,768 (39.36)	34,406 (41.70)	30,837 (39.41)	34,339 (41.65)	30,868 (39.48)

PM_2.5 fine particulate matter with diameter < 2.5 μm, PM₁₀ particulate matter with diameter < 10 μm, NO₂ nitrogen dioxide, NOx nitrogen oxides, BMI body mass index

Associations between air pollution and AAO of chronic diseases

For almost all diseases, participants with higher air pollution exposure had an earlier age of chronic disease onset (Supplementary Fig. 1). More specifically, linear associations were found for six representative diseases of each system: hypertension, COPD, schizophrenia, chronic liver disease, prostate cancer, and diabetes. There was a consistently inverse association of four air pollutant concentrations with age at onset of these diseases, with the coefficients ranging − 0.04 to − 0.33.

In the accelerated failure time (AFT) model, higher levels of air pollutants were associated with earlier onset of several chronic diseases (Fig. 1 and Supplementary Fig. 2–4). The TRs and 95% CI for most chronic diseases were less than 1 for each IQR increase in pollutant levels, indicating that air pollution accelerates the onset of the respective chronic diseases. Specifically, within the cardiovascular system, air pollution was linked to earlier onset of hypertension, atrial fibrillation, heart failure, heart attack, arterial embolism and thrombosis, phlebitis and thrombophlebitis, and nonspecific lymphadenitis. For hypertension, each IQR increase in air pollution was associated with a decrease in time to onset ranging from 0.6% for NO₂ (Adj. TR = 0.994, 95% CI = 0.993, 0.995) to 1% for PM_2.5 (Adj. TR = 0.990, 95% CI = 0.989, 0.991).

Fig. 1.

The time ratios (TRs) and 95% confidence interval (CI) for the associations between PM_2.5 exposure and occurrence timing of the chronic diseases. The TRs were associated with per IQR increase in air pollution. Accelerated failure time models were adjusted for age, sex, ethnicity education level, body mass index, smoking status, alcohol consumption, physical activity, and Townsend deprivation index (included as quantiles). Red circle indicates the reference line of TR = 1. Asterisks denote the significance level of associations (*P < 0.05, **P < 0.01, and ***P < 0.001). PM_2.5, fine particulate matter with diameter < 2.5 μm

A significant association was also observed between air pollution and earlier onset of chronic respiratory diseases, particularly vasomotor and allergic rhinitis. TRs were 0.974 (95% CI—0.971, 0.977) for PM₁₀, 0.977 (95% CI—0.974, 0.980) for PM₁₀, 0.987 (95% CI—0.985, 0.989) for NO₂, and 0.986 (95% CI—0.984, 0.989) for NO_x. Additionally, air pollution significantly accelerated the onset of neurological and psychiatric disorders, especially for schizophrenia, dystonia, and myasthenia gravis. The onset of schizophrenia was estimated to be accelerated by 2.4% for PM_2.5 (TR = 0.976), 1.8% for PM₁₀ (TR = 0.982), 3.4% for NO₂ (TR = 0.966), and 3.8% for NO_x (TR = 0.962). For diseases of other systems, air pollution also contributed to earlier disease onset. However, its impact on age of cancer onset was relatively small, primarily mainly affecting prostate cancer and leukemia.

Linear exposure–response relationships were observed between air pollution concentration and standardized age of onset for hypertension, COPD, schizophrenia, chronic liver disease, prostate cancer, and diabetes (Fig. 2). As the concentration of air pollution exposure increased, the age of onset for each of these chronic diseases showed a significant downward trend.

Fig. 2.

The dose–response relationship between air pollution and occurrence timing of for six representative diseases from each system. Models were adjusted for age, sex, ethnicity, education level, body mass index, smoking status, alcohol consumption, physical activity, and Townsend deprivation index (included as quantiles). AAO, age at onset; COPD, chronic obstructive pulmonary disease; PM_2.5, fine particulate matter with diameter < 2.5 μm; PM₁₀, particulate matter with diameter < 10 μm; NO₂, nitrogen dioxide; NOx, nitrogen oxides

Estimation of the accelerated age of onset attributable to air pollution overexposure

As shown in Fig. 3 and Supplementary Fig. 5–7, the total accelerated AAO of the 78 chronic conditions attributable to PM_2.5 overexposure (i.e., exceeding the WHO guideline of 5 μg/m³) was estimated to reach 539,320 person-years. Hypertension was the largest contributor, accounting for 18.10% of this total, followed by asthma (6.03%, 32,543 person-years), diabetes (5.39%, 29,071 person-years), and thyroid disorders (5.24%, 28,298 person-years). Similarly, the total accelerated AAO attributable to PM₁₀ overexposure was estimated at 160,105 person-years with hypertension, diabetes, and asthma as the top three contributors, accounting for 17.38%, 6.29%, and 5.48%, respectively. For NO₂ and NO_x, the estimated AAO attributable to overexposure exceeded 1.3 billion person-years, with hypertension remaining the greatest contributor (11.96% and 11.47%).

Fig. 3.

The total AAO attributable to PM_2.5 overexposure according the WHO guideline 2021. AAO, accelerated age at onset; PM_2.5, fine particulate matter with diameter < 2.5 μm. The total accelerated AAO of 78 chronic diseases attributable to PM_2.5 overexposure was estimated to reach 539,320 person-years. The total accelerated AAO of cardiovascular diseases, respiratory diseases, neurological/psychological disorders, digestive diseases, cancers, and other diseases were 148,545, 61,487, 96,392, 45,424, 17,810, and 169,662 person-years, respectively

As illustrated in Fig. 4 and Supplementary Fig. 8–10, the average AAO due to PM_2.5 overexposure was as high as 2.64 years (95% CI—2.35, 2.94), with vasomotor and allergic rhinitis being the most affected conditions, followed by nonspecific lymphadenitis (2.53 years), schizophrenia (2.45 years), and dystonia (2.22 years). For PM₁₀, the average AAOs attributable to overexposure were all less than 1 year, with nonspecific lymphadenitis, vasomotor and allergic rhinitis, and schizophrenia ranking as the top three most affected conditions. Additionally, AAO was notably higher when NO₂ and NO_x levels approached the WHO standard, particularly for schizophrenia and nonspecific lymphadenitis, both of which exceeded 8 years.

Fig. 4.

The average accelerated AAO attributable to PM_2.5 overexposure according the WHO guideline 2021. AAO, age at onset; PM_2.5, fine particulate matter with diameter < 2.5 μm

Prediction of the difference in AAO for high and low levels of pollutants

The predicted differences in the AAO of chronic diseases between participants exposed to high and low levels of pollutants revealed a significant impact of air pollution on earlier disease onset. High levels of air pollution were estimated to accelerate disease onset by 0.01 (PM_2.5 for phlebitis and thrombophlebitis) to 5.01 (NO₂ for schizophrenia) years (Fig. 5 and Supplementary Fig. 11–13). Air pollution significantly lowered AAO for neurological and psychiatric disorders; high PM_2.5 was associated with earlier onset of dystonia (shortened by 3.28 years), myasthenia gravis (1.95 years), malnutrition (1.62 years), disorder of trigeminal nerve (1.35 years), fracture (1.31 years), hypertension (1.24 years), hydrocephalus (1.13 years), and urticaria (1.01 years).

Fig. 5.

The prediction of the difference of AAO for high and low levels of PM_2.5. AAO, age at onset; PM_2.5, fine particulate matter with diameter < 2.5 μm

PM₁₀ and PM_2.5 exhibited similar effects, with high PM₁₀ exposure contributing to earlier onset of vasomotor and allergic rhinitis (1.65 years) and nasal polyp (1.00 years). Consistent findings were observed for NO₂ and NO_x exposure, with schizophrenia, esophageal varices, and dystonia being the top three affected conditions. Specifically, compared to low air pollution levels, high level of air pollution levels shortened age on onset of these conditions respectively by 5.00, 3.74, and 2.80 years for NO₂, and 4.82, 3.83, and 2.20 years for NO_x.

Stratified analyses and sensitivity analyses

Stratified analyses revealed consistent associations between air pollution and the onset age of the chronic diseases across sex and age groups (Supplementary Fig. 14 and 15). Sensitivity analyses further demonstrated that the associations between long-term exposure to air pollution and age of disease onset remained robust. Excluding participants diagnosed with diabetes within the first 3 years of follow-up did not substantially alter the results (Supplementary Fig. 16). This consistency was also observed across exposure windows of 1 year, 2 years, and 3 years prior to baseline (Supplementary Table 3–5).

Air pollution and incident risk of multiple chronic diseases

The associations between long-term exposure to pollutants and the risk of multiple chronic diseases are shown in Supplementary Fig. 17. After adjustment, air pollution exposure was associated with a higher risk of most chronic diseases, with particularly strong effects observed for respiratory diseases, neurological conditions, and psychological disorders. Schizophrenia was the most affected, with a hazard ratio (HR) of 1.25. O₂ and NO_x exposure exhibited similar effect patterns, showing significant associations with nearly all chronic diseases studied, except cancer. Specifically, PM_2.5 and PM₁₀ exposure demonstrated the strongest effects on the risk of vasomotor and allergic rhinitis, with HRs ranging from 1.492 to 1.525 for PM_2.5 and from 1.430 to 1.504 for PM₁₀ over 1, 2, 3, and 4 years of exposure.

Air pollution contributed significantly to the population attributable fractions (PAFs) across all chronic conditions (Supplementary Fig. 18–21). Specifically, PM_2.5 and PM₁₀ were major contributors to vasomotor and allergic rhinitis, with PAFs of 20.9% and 20.3% respectively. Additionally, PM_2.5 and PM₁₀ accounted for more than 10% of the PAFs for several neurological diseases, including schizophrenia, polyneuropathies, and dystonia. For other chronic conditions, the individual contributions were less than 10% of the PAFs. NO₂ and NO_x also played a critical role in the PAFs of chronic diseases across various systems. Notably, schizophrenia and nonspecific lymphadenitis were mostly affected by NO₂ and NO_x, with PAFs of 19.4% and 18.42% for schizophrenia and 18.7% and 20.4% for nonspecific lymphadenitis, respectively.

Discussion

Leveraging data from the UK Biobank cohort study, this study found that exposure to high levels of air pollution was associated with an earlier onset of 48 of 78 chronic diseases, including hypertension, stroke, COPD, diabetes, and dementia. Notably, high air pollution exposure significantly accelerated the onset of neurological and psychiatric disorders, such as dystonia and myasthenia gravis, by approximately 2 ~ 5 years. Schizophrenia was similarly impacted, with the age of onset reduced by approximately 2.4 ~ 3.8%. Overall, a substantial total AAO of the 78 chronic conditions was attributable to air pollution overexposure, with hypertension, diabetes, and asthma emerging as the top three contributors.

The association between higher level of air pollution and the risk of multiple chronic diseases is well-established [10, 13, 29–31]. However, few studies have explored the impact of air pollution on the accelerated AAO of chronic diseases. A study from southern Italy found a significant association between yearly PM₁₀ and ozone level with mean age at type 1 diabetes onset [32]. Similarly, a cohort study in Hunan, China, found that long-term PM_2.5 exposure was linked to early-onset diabetes incidence [33]. A 2024 study reported an earlier onset of lung cancer among individuals living within 600 m of an iron foundry, with onset 3.2–7.7 years earlier in men and 11.7–16.8 years earlier in women [34]. Our analysis advances prior research by quantifying the extent to which air pollution levels are associated with differences in age at onset for multiple chronic diseases. This novel contribution offers critical insights into how environmental factors may accelerate the biological aging process and precipitate the earlier manifestation of chronic conditions.

Findings from this study demonstrate the significant impact of air pollution on the early onset of neurological and psychiatric disorders. Specifically, air pollution was associated with earlier onset of schizophrenia, dystonia, and myasthenia gravis, advancing their onset by approximately 2–5 years for each interquartile range (IQR) increase in air pollutant concentrations. A further 2.4 to 3.8% reduction in the age of onset for schizophrenia associated with air pollution exposure aligns with the hypothesis that air pollutants exacerbate oxidative stress and inflammation, thereby contributing to earlier disease onset [35]. This is particularly alarming given the rising prevalence of neurodegenerative diseases and substantial burden they impose on healthcare systems and society. The Global Burden of Disease Study on neurological diseases reported that, in 2021, 3.4 billion people worldwide were affected by neurological conditions, accounting for 443 million years of healthy life lost, surpassing cardiovascular diseases as the leading contributor to global disease burden [36]. Given the pervasive nature of air pollution and its significant public health implications, it is crucial to implement comprehensive strategies aimed at reducing pollutant emissions and mitigating exposure risks to slow the onset of neurological diseases.

In this analysis, chronic diseases such as hypertension, asthma, diabetes, COPD, obesity, and thyroid disorders still accounted for the majority of total accelerated AAO. This is likely attributable to the high incidence of these conditions, making them key targets for prevention and control efforts. Although these conditions are already a common target of intervention, traditional public health efforts have historically focused on behavioral change, such as improving diet and physical activity, which is notoriously challenging to achieve and sustain. Shifting the focus to environmental-level interventions, such as policies aimed at reducing particulate pollution, may offer greater population-level-impact. Additionally, evidence linking air pollution exposure to earlier age of onset presents a new challenge for prevention and control strategies. Historically, younger populations have not been the primary target for intervention for conditions such as hypertension, diabetes, and COPD. However, as these diseases increasingly affect younger demographics, addressing this shift will be essential. Progress in managing the growing burden of chronic disease will require effective public health strategies that include environmental interventions and target younger populations, who are increasingly affected [37]. In these ways, mitigating air pollution could delay disease onset and reduce associated healthcare demands, ultimately promoting better population health outcomes.

This study highlighted the significant public health implications of long-term exposure to air pollution, demonstrating its strong association with an increased risk and accelerated onset of various chronic diseases. The findings suggested that exposure to pollutants not only elevated the risk of these conditions but also brought forward their onset by several years, thereby extending the duration of morbidity and increasing the healthcare burden. This study has profound implications for health policy, underscoring the need for stricter air quality regulations, urban planning reforms, and targeted interventions to reduce pollution levels and protect vulnerable populations. Moreover, there is a need for increased public awareness about the health risks associated with air pollution, which can motivate advocacy for such policy change as well and adoption of prevention measures at the individual level. Overall, this research emphasizes the urgent need for comprehensive, evidence-based strategies to mitigate the impact of air pollution on health, reduce the early onset of chronic diseases, and promote health equity across diverse populations.

The potential mechanisms underlying the association between air pollution and the earlier onset of chronic diseases remain uncertain. However, oxidative stress and inflammation are well-recognized pathways through which pollutants such as PM_2.5, PM₁₀, NO₂, and NO_x induce systemic damage [38–40]. These pollutants might lodge deep in the respiratory system and enter the bloodstream, triggering systemic inflammatory responses and oxidative damage in multiple organs, ultimately accelerating disease onset [41, 42]. Furthermore, socioeconomic factors linked to higher exposure of air pollution, such as living in poverty, decreased access to healthcare, and higher prevalence of smoking and poor diet, may also contribute to the earlier onset of the chronic diseases studied here [43, 44]. Together, these biological and social determinants highlight the need for comprehensive public health strategies that not only aim to reduce air pollution but also address these underlying social determinants of health. Such an approach would ensure a more holistic response to mitigating the impact of air pollution on population health.

This study has several strengths, including its large sample size, extensive follow-up period, and use of advanced statistical analyses to assess the impact of air pollution on the age at onset of a wide range of chronic diseases. However, observational nature of the study limits the ability to infer causality, and residual confounding cannot be entirely ruled out. Additionally, the exposure assessment was based on residential addresses, which may not fully capture individual-level exposures due to variations in time spent indoors and outdoors, occupational exposures, and other factors. Furthermore, the data were obtained from a study conducted in the UK, where the combination of pollutants studied may differ those in other regions worldwide. As such, findings may not be generalizable to other settings. The effects on AAO could be even larger in countries such as India and China where air quality issues are well-documented. Future research conducted in diverse geographic regions may reveal policy nuances needed in specific countries. Additionally, future studies should focus on elucidating the biological mechanisms underlying the associations observed in this study. Understanding these mechanisms would strengthen the argument for causality and provide robust evidence to advocate for and inform policy change. Intervention studies that evaluate the impact of air quality improvements on the onset age of chronic diseases would also offer valuable insights into the potential benefits of pollution reduction strategies. Such evidence could serve as a model for other communities or countries seeking to implement effective air quality interventions and policies.

In conclusion, this study provides novel evidence that long-term exposure to ambient air pollutants is significantly associated with an increased risk and accelerated onset of various chronic diseases. These findings underscore the urgent need for effective air quality management policies and targeted interventions to reduce exposure to harmful pollutants. Such steps are essential not only to mitigate their detrimental effects on public health but also to delay the onset of chronic diseases, ultimately reducing the societal and economic burden they impose.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

HL supervised the entire project and design the work. HL and FT accessed and verified the data. FT contributed to the data analysis, data interpretation, and writing of the manuscripts. SW, ZQ, JZ, YW, KH, LA, and TB contributed to the discussion and revised the manuscript. All authors reviewed or revised the manuscript and approved the final draft for submission.

Data availability

Researchers can apply to use the UK Biobank resource for health-related research that is in the public interest (https://www.ukbiobank.ac.uk/enable-your-research/register).

Declarations

Competing interests

The authors declare no competing interests.

Disclosures

None.