The mediation effect of sleep quality on the association between ambient air pollution and type 2 diabetes among UK adults: a population-based cohort study – BMC Public Health

Report on a Population-Based Study of Environmental and Health Determinants

This report details the methodology of a large-scale prospective cohort study investigating the interplay between environmental exposures, lifestyle factors, and the incidence of non-communicable diseases. The study’s framework and objectives are directly aligned with several United Nations Sustainable Development Goals (SDGs), providing critical evidence for public health policy.

• SDG 3 (Good Health and Well-being): By focusing on the incidence of Type 2 Diabetes (T2D), the study directly addresses Target 3.4, which aims to reduce premature mortality from non-communicable diseases.
• SDG 11 (Sustainable Cities and Communities): The assessment of air pollution as a primary exposure variable contributes to Target 11.6, which seeks to reduce the adverse per capita environmental impact of cities, particularly concerning air quality.
• SDG 10 (Reduced Inequalities): The inclusion of socioeconomic and demographic covariates, such as the Townsend Deprivation Index and ethnicity, allows for an analysis of how health outcomes are distributed across different population segments, supporting efforts to reduce health inequalities.

Methodology

Study Population and Data Linkage

The investigation utilized data from the UK Biobank, a major population-based cohort study that supports research into a wide range of complex diseases, aligning with SDG 17 (Partnerships for the Goals) through its open-access data-sharing model.

• Participants: The initial cohort included over 500,000 individuals aged 40–69, recruited between 2006 and 2010 from 22 assessment centers across Scotland, England, and Wales.
• Data Sources: Health outcomes were ascertained via record linkage to national databases, including hospital inpatient records and the National Death and Cancer Registry (NDCR).
• Final Cohort: After applying exclusion criteria (e.g., pre-existing T2D, incomplete data), the final analytical cohort comprised 226,188 participants.

Outcome and Exposure Assessment

The study’s primary outcome was the incidence of T2D, a critical health indicator for monitoring progress toward SDG 3. Environmental exposure assessment focused on air quality, a key component of SDG 11.

1. Primary Outcome: Incident T2D was identified using hospital inpatient data, coded as E11 under the International Classification of Diseases, Tenth Edition (ICD-10).
2. Exposure Assessment: Annual mean concentrations of air pollutants were estimated using ESCAPE Land Use Regression (LUR) models at a high spatial resolution (100 × 100 m grid). This detailed environmental monitoring provides crucial data for creating healthier urban environments.
3. Pollutants Measured:
   • Particulate Matter (PM₂.₅, PM₁₀)
   • Nitrogen Dioxide (NO₂)
   • Nitrogen Oxides (NOₓ)
4. Exposure Assignment: Each participant’s baseline residential address was geocoded and linked to the corresponding grid cell to assign an individual-level exposure value based on mean annual pollutant concentrations.

Sleep Quality and Covariate Measurement

A comprehensive assessment of sleep quality and potential confounding variables was conducted to ensure a robust analysis of health determinants, reflecting the multifaceted nature of achieving SDG 3 and addressing inequalities as per SDG 10.

• Sleep Quality Score: A composite sleep score (ranging from 0 to 7) was derived from seven self-reported behaviors: sleep duration, chronotype, daytime sleepiness, insomnia, snoring, nap frequency, and difficulty waking. Higher scores indicated poorer sleep quality.
• Covariates: A wide range of potential covariates were identified using directed acyclic graphs (DAGs) to control for confounding. These included:
  • Demographic: Age, sex, ethnicity, assessment center.
  • Socioeconomic: Educational level and the Townsend Deprivation Index, which measures material deprivation.
  • Lifestyle: Smoking status, alcohol consumption, fruit and vegetable intake, physical activity (MET minutes), and Body Mass Index (BMI).

Analytical Approach

Statistical and Risk Modeling

Advanced statistical models were employed to quantify the association between air pollution and T2D incidence, providing the evidence base needed to inform policies aimed at achieving health and environmental SDGs.

1. Hazard Models: Cox proportional hazards and Aalen additive hazards models were used to estimate hazard ratios (HR) and parameter estimates (β) for T2D risk per 10 µg/m³ increase in each air pollutant.
2. Model Adjustment: Two models were constructed: Model 1 adjusted for age and sex, while the fully adjusted Model 2 included educational level, ethnicity, Townsend Deprivation Index, and assessment center.
3. Non-Linearity Assessment: Restricted cubic splines (RCS) were used in adjusted Cox models to investigate potential non-linear relationships between air pollution exposure and T2D incidence.

Causal Mediation and Sensitivity Analyses

To enhance the validity and reliability of the findings, causal mediation and a series of sensitivity analyses were performed. This rigorous approach strengthens the study’s contribution to evidence-based strategies for sustainable development.

• Causal Mediation Analysis: This analysis was conducted to investigate the mediating role of sleep quality in the relationship between air pollution and T2D, estimating the natural direct and indirect effects.
• Sensitivity Analyses: Several analyses were performed to test the robustness of the results:
  1. Multiple imputation for missing covariate data.
  2. Exclusion of T2D cases diagnosed within the first three years of follow-up to minimize reverse causality.
  3. Restriction of the analysis to participants with stable residence (>3 years).
  4. Re-analysis using air pollution data from the UK Department for Environment, Food and Rural Affairs (DEFRA).
  5. Application of two-pollutant models to assess pollutant-specific effects.
  6. Use of a Fine and Gray competing risk regression model to account for mortality.

Analysis of Sustainable Development Goals in the Article

1. Which SDGs are addressed or connected to the issues highlighted in the article?

The article addresses and connects to several Sustainable Development Goals (SDGs) through its investigation into the links between environmental factors, health outcomes, and socioeconomic determinants.

• SDG 3: Good Health and Well-being

  This is the most central SDG to the article. The study’s primary objective is to investigate health outcomes, specifically the “incidence of T2D (Type 2 Diabetes),” which is a major non-communicable disease (NCD). The entire research framework, from participant selection to outcome identification, is focused on understanding factors that contribute to poor health and disease, aligning directly with the goal of ensuring healthy lives and promoting well-being.

• SDG 11: Sustainable Cities and Communities

  The article is directly relevant to this goal because it focuses on a key aspect of urban environmental quality: air pollution. The study assesses exposure to various air pollutants, including “particulate matter (PM₂.₅, PM₁₀), NO₂, and NOₓ,” which are significant environmental hazards, particularly in urban and industrial areas. By examining how these pollutants, measured in “urban, suburban, and rural settings,” affect public health, the article contributes to the knowledge base needed to make cities and human settlements safer and more sustainable.

• SDG 10: Reduced Inequalities

  This SDG is addressed through the study’s inclusion of socioeconomic and demographic factors as potential covariates. The analysis incorporates variables such as “ethnicity, socioeconomic status, [and] educational level.” Specifically, the use of the “Townsend Deprivation Index,” which “evaluates deprivation based on the combined scores of four census variables: unemployment rate, car ownership rate, home ownership rate, and family overcrowding,” shows a clear effort to understand how health outcomes are distributed across different socioeconomic groups. This analysis helps to highlight health inequalities that may exist within the population.

2. What specific targets under those SDGs can be identified based on the article’s content?

Based on the article’s focus, specific targets within the identified SDGs can be pinpointed.

1. Target 3.4 (under SDG 3)

   “By 2030, reduce by one third premature mortality from non-communicable diseases through prevention and treatment and promote mental health and well-being.”

   The article directly supports this target by investigating the environmental and behavioral risk factors for Type 2 Diabetes, a prominent non-communicable disease. The study’s primary outcome is the “incidence of T2D,” and by identifying a relationship between air pollution, sleep quality, and the development of T2D, it provides crucial information for prevention strategies.

2. Target 11.6 (under SDG 11)

   “By 2030, reduce the adverse per capita environmental impact of cities, including by paying special attention to air quality and municipal and other waste management.”

   This target is addressed by the study’s core methodology, which involves an “Exposure assessment” of air quality. The research utilizes Land Use Regression models to “estimate annual mean concentrations of particulate matter (PM₂.₅, PM₁₀), NO₂, and NOₓ at a high spatial resolution.” This focus on measuring and analyzing the health impacts of urban air quality is central to Target 11.6.

3. Target 10.2 (under SDG 10)

   “By 2030, empower and promote the social, economic and political inclusion of all, irrespective of age, sex, disability, race, ethnicity, origin, religion or economic status or other status.”

   The article connects to this target by analyzing health outcomes while accounting for social and economic disparities. The inclusion of covariates like “ethnicity,” “educational level,” and the “Townsend Deprivation Index” in the statistical models allows for an examination of how health risks (like developing T2D from air pollution) may disproportionately affect more deprived or marginalized populations, thereby providing evidence relevant to reducing health inequalities.

3. Are there any indicators mentioned or implied in the article that can be used to measure progress towards the identified targets?

Yes, the article mentions and uses data that directly correspond to or serve as proxies for official SDG indicators.

• Indicator for Target 3.4

  The official indicator is Indicator 3.4.1: “Mortality rate attributed to cardiovascular disease, cancer, diabetes or chronic respiratory disease.”

  The article provides a direct measure related to this by focusing on the “incidence of T2D.” While incidence is not the same as mortality, it is a critical precursor and a direct measure of the burden of the disease within the population. Tracking the incidence of diabetes is fundamental to understanding and eventually reducing mortality from it.

• Indicator for Target 11.6

  The official indicator is Indicator 11.6.2: “Annual mean levels of fine particulate matter (e.g. PM2.5 and PM10) in cities (population weighted).”

  The article explicitly measures this. The “Exposure assessment” section states that the study was “utilized to estimate annual mean concentrations of particulate matter (PM₂.₅, PM₁₀), NO₂, and NOₓ.” This data is precisely what is required for Indicator 11.6.2, making the study’s methodology directly applicable to monitoring this SDG target.

• Indicator for Target 10.2

  The official indicator is Indicator 10.2.1: “Proportion of people living below 50 per cent of median income, by age, sex and persons with disabilities.”

  While the article does not use this exact indicator, it employs a powerful and relevant proxy: the “Townsend Deprivation Index.” This index is described as a measure of socioeconomic status that evaluates deprivation based on unemployment, car ownership, home ownership, and overcrowding. It serves as a comprehensive indicator of economic and social inequality within the UK context. By analyzing health outcomes stratified by this index, the study provides insight into health disparities linked to socioeconomic status, which is the core principle of Target 10.2.

4. Summary Table of SDGs, Targets, and Indicators

Source: bmcpublichealth.biomedcentral.com