Low-level pollution - health impacts: evidence review

This review explores the existing evidence on the health effects associated with low-level pollution in countries that have levels of ambient air pollution similar to Scotland.

2 Methodology

2.1 Literature search

A robust, rapid review of the available scientific literature was undertaken using the Scopus database. Due to the time and resource constraints of the study only Scopus was used, however Scopus is a relatively comprehensive database and additional searches were completed to ensure all relevant evidence was retrieved. Prior to the main literature search, a preliminary search was conducted using research databases and search engines. The purpose of this was to identify any important evidence that could frame the search strategy and terms. After the main literature search, supplementary searches using Google were conducted to ensure all relevant literature was obtained. Although the review is not a full systematic review, the search was conducted in a systematic way and using appropriate evaluation of the quality of sources to reflect the balance of evidence.

2.1.1 Search terms

The preliminary literature search identified a mapping review by Dominski et al. (2021) ‘Effects of air pollution on health: A mapping review of systematic reviews and meta-analyses’, herein referred to as the Dominski review. The Dominski review identified 240 evidence reviews that analysed the effects of air pollution on human health. The Dominski review is of important relevance to the research question and collates all the high-quality evidence synthesis papers published on or before 18 June 2020. However, to fully answer the research question, we have adapted the search terms in the Dominski review to collate evidence published after this date, as well as data that could be comparable to Scotland.

Table 1 provides a breakdown of the search terms used. Search terms [1] [2] and [4a] were based upon the terms used by the Dominski review (Dominski et al., 2021). The additional health search terms [3], were identified to encompass a wider range of both physical and mental health outcomes from air pollution. The search terms used to exclude weaker study designs [4b] were ascertained following in-depth discussions using the project team’s subject area and methodological expertise and verified with Scottish Government. The timescale of the study [5] was designed to collect any new evidence, not published at the time of the Dominski review and the location [6] included countries comparable to Scotland, defined by population-weighted annual mean PM2.5 data, from the Health Effects Institute’s 2020 State of Global Air (Health Effects Institute, 2020). PM2.5 was used as this pollutant is most commonly associated with health effects (UKHSA, 2022). To define countries as comparable to Scotland, data on the population-weighted mean exposure in either 2017, 2018 or 2019 (pre-COVID-19) was collected with a threshold of 12 µg/m3, as the United Kingdom (including Scotland) mean is approximately 10.2 µg/m3 (based on the State of Global Air 2017-2019 data). Finally due to the time implications of translation, only studies published in English language [7] were included. References and documents were stored and classified using Zotero reference management software.

2.1.2 Search limits

The search was limited to English language studies published between 1 January 2020 and 1 January 2024. Although the last search date in the review by the Dominski review is 18 June 2020, the search strategy used in this study was from 1 January 2020 to ensure all relevant papers were collated, including those that may have been accepted in early 2020 but not published until after 18 June 2020, due to the timescales associated with academic publishing. An end date of 1 January 2024 was used to capture any pre-dated early release publications.

Table 1: Search terms and combinations
Search terms Descriptors Justification Reference
1. Air pollution “air pollution” OR “air pollutant*” OR “air quality” OR “particulate matter” OR “PM10” OR “PM2.5” OR “carbon monoxide” OR “carbon dioxide” OR “ozone” OR “nitrogen dioxide” OR “sulfur dioxide” OR “traffic-related air pollution” Based upon mapping review of systematic reviews/meta-analyses (Dominski review) Dominski et al. (2021)
2. Physical health “health” OR “public health” OR “cardiovascular disease” OR “ischemic heart disease” OR “cancer” OR “respiratory disease” OR “stroke” OR “cerebrovascular disease” OR “morbidit*” Based upon mapping review of systematic reviews/meta-analyses (Dominski review). Dominski et al. (2021)
3. Mental health and cognition OR "anxiety" OR "depression" "cognit*" OR "memory" OR "cerebr*” OR “mental*" OR "declin*" or "deteriorat*" OR "degenerat*" OR "Neurodegenerative" OR "disorder" OR "neurological" OR "psych*" Inclusion of wider health effects following discussion with Scottish government. Zundel et al. (2022) Bui et al. (2018) Sun et al. (2015)
4a. Study type “systematic review” OR “meta-analysis” OR “meta-analyses” Based upon mapping review of systematic reviews/meta-analyses. Dominski et al. (2021)
4b. Study type AND NOT “case report” AND NOT “case-report” AND NOT "cross-sectional" AND NOT "cross sectional"
5. Timescale 01/01/2020 – 01/01/2024 To collect new data, not encompassed by the mapping review of systematic reviews/meta-analyses. Dominski et al. (2021)
6. Location AND "American Samoa" OR "Andorra" OR "Australia" OR "Austria" OR "Bermuda" OR "Brazil" OR "Brunei" OR "Darussalam " OR "Canada" OR "Cook Islands" OR "Denmark" OR "Estonia" OR "Fiji" OR "Finland" OR "France" OR "Germany" OR "Greenland" OR "Guam" OR "Iceland" OR "Kiribati" OR "Latvia" OR "Lithuania" OR "Luxembourg" OR "Maldives" OR "Marshall Islands" OR "Micronesia" OR "Monaco" OR "Nauru" OR "Netherlands" OR "New Zealand" OR "Niue" OR "Northern Mariana " OR "Norway" OR "Palau" OR "Portugal" OR "Puerto Rico" OR "Russian Federation" OR "Saint Kitts and Nevis" OR "Samoa" OR "San Marino" OR "Spain" OR "Sweden" OR "Switzerland" OR "Tokelau" OR "Tonga" OR "Tuvalu" OR "United Kingdom" OR "United States of America" OR "United States Virgin Islands" OR "Uruguay" OR "England" OR "Scotland" OR "Northern Ireland" OR "Wales" OR "USA" OR "Australasia" OR "Caribbean" OR "Eastern Europe" OR "High-income Asia Pacific" OR "High-income North America" OR "Oceania" OR "Southeast Asia" OR "Southern Latin America" OR "Tropical Latin America" OR "Western Europe" Countries representative of ‘low-level pollution’. This is filtered from the 2020nState of Global Air using PM2.5 (which is most related to health effects) and setting a threshold of 12 µg/m3 as the population-weighted exposure mean in any of the years 2017, 2018 or 2019 (pre-COVID-19). The UK mean is approximately 10.2 µg/m3 for comparison. Independent Scottish data not available in this dataset. Health Effects Institute (2020)
7. Language English


Email: andrew.taylor2@gov.scot

Back to top