README.md

In aqli-epic/epi.meta.analysis: An Amazing Shiny App

Epidemiological Studies: Meta Analysis

This analysis is our ongoing attempt to capture all epidemiological research studies (>1000 people; max sample size: 1380 million), that are relatively long term (> 1 year(s), max study duration: 36 years), with unique cohorts (by unique we mean that even if a given cohort is studied by different groups over different years/different sample size, it’ll be counted as separate cohorts, instead of 1) and measure the impact of ambient PM2.5, PM10, TSP, Ultra Fine Particulate Matter on Mortality (cause-specific, all cause, premature)/Life Expectancy published between 1993 and present*, findable in available peer-reviewed literature. Hereafter, we refer to these studies as “AQ epi studies” for short. As of now, this meta-analysis analyzes 84 AQ epi studies. This analysis will be continually updated to incorporate new AQ epi studies.

We are seeking to make this analysis as current, complete and error-free as possible and view it as a continual work in progress. We would appreciate the air quality community’s comments, corrections, and suggestions. Please contact [e-mail] or leave a comment in this GitHub repository/directly leave comments in the *analysis dataset* (more on this below).

Major Takeaways

• 53.57 percent of all AQ epi studies (45 studies) included in this analysis included populations in either the US or Canada or both.

• Of the total number of times a given continent’s population has been included in any given AQ epi study, North America dominates the rest of the continents and has been included 53.57 percent of times (all due to studies in the US or Canada). Closely following North America, populations in Asia and Europe have been included in AQ epi studies 28.57 and 17.86 percent of times respectively. Africa, South America, Oceania have seen 0 long term (>= 1 year), large (>=1000 people in the sample) AQ epi study.

• There have been no large (>= 1 year) or long (>=1000 people in the sample) PM epi studies that were performed in Africa, South America, Oceania, with a combined population of 1.84 billion people , 23.8 percent of the Earth’s total human population.

• In total, 7 “really” long term studies (> 25 years) have been performed. That is, 8.33 percent of the total number of studies.

• All of the following continents have seen at least 1 really long term study: North America, Europe.

Purpose

The purpose of this analysis is to understand the landscape of epidemiological research on the relationship between PM2.5, PM10, TSP, Ultrafine Particulate Matter and Mortality (all types, as specified above)/Life Expectancy and to surface demographic, geographic, or other trends that may exist in the current state of literature. While the overall arc of the relationship between these pollutants and human health is clear enough to take action, understanding such trends can help the field reflect on itself, take stock of any biases or gaps – and point toward future research and policy opportunities.

Why do epidemiological studies on air pollution and mortality matter?

While global estimates of air pollution’s toll on public health vary, they all point in the same direction: air pollution poses one of the largest health risks on the planet to humans [Paper1, Paper2, Paper3, Paper4]. Epidemiological studies on air pollution and mortality help us understand the burden of air pollution on human health at global, national and regional levels. According to Vahlsing and Smith (2012), these sorts of studies can also help countries take policy action, pushing forward and shaping national-level ambient air quality standards.

The burden of air pollution across the world is also not uniform. While 98.3 percent of the world population is out of compliance with the latest WHO annual PM2.5 guideline of 5 µg/m³, there is huge variation in the quality of air one breathes.

Results

PM2.5 concentration range and the Global Population Distribution

• In total 84 AQ epi studies were included in the final analysis dataset. Of these, 73 were PM2.5 specific. Others (11 studies) are multi-pollutant studies.

• 12.7 percent of the world population, or 983.9 million people, live in areas where the annual average PM2.5 pollution is greater than 50 µg/m³. But, only 9.6 percent (7 PM2.5 specific studies) of the total PM2.5 studies, have been performed in these highly polluted parts of the world. These highly polluted areas are areas where the average PM2.5 pollution is at least 10 times the WHO PM2.5 safe guideline of 5 µg/m³.

• Approximately 5.9 percent of the world population (458.3 million people) live in the most severely polluted parts of the world, where annual average PM2.5 pollution is upwards of 75 µg/m³ (at least 15 times the WHO safe guideline). In these most severely polluted parts of the world, 0 PM2.5 AQ epi studies have been performed .

• Most of the PM2.5 AQ epi studies (67.1 percent of the total number of PM2.5 studies, or 49 studies) performed so far, are concentrated in areas where the average PM2.5 concentration is in the 0-25 µg/m³ range. People living in these areas (58.3 percent of the world population) are breathing air that is much less polluted relative to the people living in the most polluted parts of the world (as seen above). But, even in the 0-25 µg/m³ bucket, anyone living above 5 µg/m³, is out of compliance with the WHO PM2.5 guideline.

Geographic Distribution of Studies

• 53.6 percent of all AQ epi studies (45 studies) included in this analysis included populations in either the US or Canada or both.

• USA, Canada and Europe (or some combination of them) were focus countries in 71.4 percent of all AQ epi studies (60 studies). Other countrie(s) that have been included in an AQ epi study: China, Japan, Hong Kong, Taiwan, India, Indonesia, South Korea.

• Of the total number of times a given continent’s population has been included in any given AQ epi study, North America dominates the rest of the continents and has been included 53.6 percent of times (all due to studies in the US or Canada). Closely following North America, populations in Asia and Europe have been included in AQ epi studies 28.6 and 17.9 percent of times respectively. Africa, South America, Oceania have seen 0 long term (> 1 year), large (>1000 people in the sample) AQ epi study.

AQ Epi studies over time

• In most of the 90’s and early 2000’s, the rate of AQ epi studies publishing was around 1 to 2 studies per annum on average.

• Post 2009, there has been a noticeable increase in the overall volume of AQ epi studies published.

Distribution of duration of study by continent

• In total, 7 “really” long term studies (> 25 years) have been performed. That is, 8.3 percent of the total number of studies.

• All of the following continents have seen at least 1 really long term study: North America, Europe.

• All of the following continents have seen more than 1 really long term study: North America.

• In all of the really long term studies (including both multi-country/continent and single-country), i.e. ones that have a study duration of > 25 years - North America was included 6 times in those studies (85.7 percent of the total number of times any given continent is included in any given study).

• Similarly, Europe has been included 1 times in such long term studies (14.3 percent of the total number of times any given continent is included in any given study).

Conclusion

All studies included in this analysis point to the same overall picture:air pollution is a serious health threat. The existing state of scientific literature on air pollution and health is clear that air pollution’s impact on health is well-established and taking action to improve a polluted environment should not be delayed in order to complete multi-year large sample (> 1000) epidemiological studies in an area, even if there has not been a prior study in that particular geography. That said, it is important for the field of air quality epidemiolgy to understand the contours of its current research landscape to most effectively identify directions for future research and deploy limited resources.

Methodology

Through a comprehensive and ongoing* literature review, we are making an attempt at creating an exhaustive public listing of all the epidemiological studies out there (that we could find) that examine the relationship between PM2.5 and Life Expectancy/Mortality.

For each study, we record data on key defining features, such as: Geography, Sample Size, Study Duration, PM2.5 exposure range, etc. Then we used this analysis dataset to carry out a meta-analysis, results of which are detailed in the Results section above.

We are seeking to make this analysis as current, complete and error-free as possible and view it as a continual work in progress. We would welcome the air quality community’s any comments, corrections, andor suggestions. Please contact [e-mail] or leave a comment in this GitHub repository.

Inclusion/Exclusion Criteria for studies

• The underlying *analysis dataset* for this meta analysis focuses only on those papers that study the link between PM2.5, PM10, TSP, Fine Particulate Matter and *Mortality/Life Expectancy*. In addition to the analysis dataset, there is a *master dataset* that expands on the analysis dataset to include other useful information. The idea behind the master dataset is to record any additional details (whether additional facts about the paper, or details on other pollutants studied in the paper) that will not be a part of the main data analysis exercise. Master dataset will also list additional papers that do not fit into the inclusion criteria.

• The analysis dataset excludes the following types of papers: meta-analysis, unpublished papers, papers studying the effects of indoor air pollution on Mortality/Life Expectancy, papers forecasting future air pollution/life expectancy/mortality. But, the master dataset lists all of these studies.

• In Multi-Country (pooled) studies, one entry (one row in the analysis dataset) is recorded for each country in papers where country level data is available. There are some pooled studies where country level data is not available, but rather data is available for a custom region (e.g. South-East Asia), such pooled studies are excluded from the analysis dataset. But, the master dataset still lists all of these additional studies for reference.

• In cases where the same cohort is studied by different research groups at different points in time/using different methods: we have included and counted all of those studies as separate unique studies in the analysis below.

• Papers studying the health effects of pollution segregated by sectors (source apportionment type studies), regions (example, Rural v/s Urban), season (Winter v/s Summer) are excluded from this analysis.

• Papers that include combined estimates of Household and Ambient Air Pollution, but not individual estimates are excluded.

• All papers studying the impact of pollutants on DALY’s (Disability Adjusted Life Years), are excluded from this analysis.

Other important points

• In papers, where the minimum PM2.5 concentration was not reported, we have assumed the lowest available percentile data available on PM2.5 concentration as the minimum concentration.

• In many papers, only one of the mean PM2.5 or PM2.5 range is reported but not both. In these cases, wherever the data is not available (whether it is mean PM2.5 or PM2.5 range) we have recorded the instances as “NA”. Apart from this, anywhere we couldn’t find data, we have recorded it as “NA”. This is one of the limitations of this analysis.

• Different papers talk about different types of mean PM2.5 values. Some report daily averages, while others may report monthly/annual averages. Also, in some cases, the averages are population weighted and in other cases they are not. But, it is not always clearly stated as to which type of average the paper is referring to (or how it was calculated). In scenarios like these, we had to make a value judgement and this is one of the limitations of this analysis.

• In cases, where mean PM2.5 is reported for both the start year and the end year(or also for sometime between a start year and the end year) of the study, we have reported the end year mean PM2.5 value.

• Papers that have used a conversion factor to convert mean PM10 concentrations to mean PM2.5 concentrations are included but, there PM2.5 values (mean, lower limit, upper limit, standard deviation) is not recorded.

• In multi-pollutant studies where different pollutants have been studied over different periods of time, we have chosen the specific time period that corresponds to the PM2.5 pollutant (in cases where we PM2.5 is not present, we have recorded the time period corresponding to one of the other pollutants).

• We have recorded one mean PM2.5 value for each paper. But, there are papers where more than one mean PM2.5 value is reported (for example, one for the male group and one for the female group). In these cases, we have picked one value from the ones that are available.

• In some papers, the exact start and/or end year of a study is unclear. In these cases, we have mostly chosen the first instance of the multiple “study duration ranges” (depending on multiple plausible interpretations of when the final follow up ended) the paper.

• In many papers, the upper limit or lower limit for age is not precisely specified. In these cases, we have recorded a NA. For example, there are papers where the upper limit category for age is 85+. In these cases, although we know the upper limit category, we still don’t know the upper limit of the age, which could be 90, 95, 100, etc.

• In places, where the sample size is specifed in terms of “number of regions” (for example, 17 districts) and not “number of people”, we have recorded a NA in the cohort size column.

• Under the “methods” column (in the master dataset), we have only mentioned a subset of methods that were used to carry out different parts of the study. Authors may have used other methods than those mentioned in our meta analysis.

• Different graphs are generated using different subsets of the analysis dataset. Most of them are PM2.5 specific (our main focus), others include all pollutants (PM2.5, PM10, TSP, Ultrafine Particles). The type of papers used to generate a given graph and other nuances for the graph in question are specified within the graph (as a note) and/or the accompanying text.

How can you (the community) help in improving this analysis?

• Add to the *analysis dataset* and and grow the epi database:

• If you know of other papers that: (a) are attempting to study the link between PM2.5 and Life Expectancy/Mortality and (b) are not included in this analysis: Please leave a comment in the *analysis dataset*, providing a link to the paper. You can also write to us at [e-mail] (with the link to the paper mentioned in the email).

• As a next step, we’ll go through your submission, and if it fits our inclusion criteria, we will update the underlying analysis dataset and re-render the entire blog, so that it represents the most up to date data and figures.

• If you are unsure (given the inclusion criteria) about whether a particular paper (that you want to post) should be posted on not, we encourage you to post it and let us worry about the inclusion/exclusion bit.

• In case you have comments on some aspects of a paper/if you find any errors, please leave a comment in the *analysis dataset* on the cell (tagging [e-mail], using **@** symbol) where the error is found or write to us at [e-mail], detailing the error and its location (cell address on the sheet).

References

Other Graphs and Interactive Summary Dashboard

To further explore these graphs and more in an interactive setup, visit the AQ Epi dashboard.

aqli-epic/epi.meta.analysis documentation built on July 2, 2023, 4:18 p.m.