Household air pollution and risk of esophageal cancer: a case-control study in Western Kenya

Lead Research Organisation: University of Liverpool
Department Name: Institute of Psychology Health & Society

Abstract

Esophageal cancer is the third most common cancer and cause of cancer death in both men and women in Kenya, with 3000+ newly diagnosed patients in 2012. This pattern is common to other East African countries and West Kenya appears to have extremely high rates of the disease. This status contrasts to much of West Africa where the cancer is extremely rare, i.e. 40 times lower. Sadly, the majority of these patients pass away within 6 months of their diagnosis, few being able to afford the limited treatment options. Additionally, up to 15% of patients are very young, in their 20s and 30s. Despite researchers having known about this unusual disease pattern since the 1950s, the risk factors for the disease had hardly been studied prior to 2014. Consequently, there are no targeted strategies to prevent the disease, and no prediction models to identify who might be at high risk of the disease, people whom may benefit from early detection.

In 2014, we - i.e. the Kenya PI Dr Diana Menya of Moi University, Eldoret - commenced the first comprehensive study of lifestyle, environmental and genetic factors for esophageal cancer in Kenya. The study was conducted at the Moi Teaching and Referral Hospital, Eldoret, which serves a catchment population stretching to the Ugandan border. In the study, patients newly diagnosed with esophageal cancer were compared to hospital patients and hospital visitors who did not have the disease. The study blood and tumour samples were also collected for genetic analyses. Results from the study have shown that alcohol and tobacco contribute to the disease in older men, but not in women or in young patients, a knowledge gap which the present study hopes to fill through a UK-Kenya collaboration between Dr Menya and Dr Daniel Pope, a household air pollution expert from the University of Liverpool. This study hypothesizes that household air pollution, from using biomass for cooking in poorly ventilated kitchens, is a large contributor to esophageal cancer in the young and in women, through traditional domestic roles associated with cooking. Household air pollution from biomass and coal use is already an established lung carcinogen, but few studies have examined its influence on esophageal cancer. However, work in an Iranian area of high esophageal cancer rates has shown that chemical compounds formed after combustion of biomass fuels are implicated in esophageal cancer risk in that setting.

In this UK-Kenya collaboration, we propose to continue the core study as previously successfully implemented, and add to it an in-depth component on household air pollution as measured in kitchens and for personal exposures during household visits to a subset of female and younger (< 40 years) participants who live within reach of the referral hospital. A detailed analysis of oesophageal cancer risk with household air pollution will be provided and finally, considering a range of lifestyle and environmental risk factors, a comprehensive report on the strategies needed for primary prevention esophageal cancer in Kenya will be developed. Rates of this disease have declined in many areas of the world - Kenya should be able to follow this trend.

Technical Summary

Squamous cell esophageal cancer is a poor prognosis common cancer in Kenya. The current aetiological model for its peculiar global geospatial variation points to the presence of a dominant environmental factor, on top of which lifestyle factors operate. In Eldoret, West Kenya, a hospital based case-control study was successfully set up by our Kenyan PI, Dr Diana Menya, in 2014 and now includes 400 case and 400 controls. It includes a comprehensive questionnaire-based assessment of household air pollution as well as other lifestyle factors. Controls are age and gender frequency-matched to cases, recruited from hospital visitors and short-term or day patients. Biospecimen collections include blood, urine and tumour samples. In the present study, we will expand this study to improve exposure assessment for suspected environmental factors, in particular of household air pollution. In addition to the on-going data collection, household visits will be made to each consenting case and control participant who indicate potential exposure to household air pollution (HAP), often through cooking activities. In the home, participants or their proxies will be provided with a microPEM 24-hour monitoring to measure PM2.5 and PM2.5-bound polycyclic aromatic hydrocarbons. Air concentration levels will be compared to WHO standards and their determinants analysed, to produce a current and lifetime HAP exposure prediction model which will then be applied to the full case-control study (650 case: 650 controls). Using this improved HAP exposure, we will investigate whether HAP is a risk factor for esophageal cancer and we will estimate corresponding population attributable fractions to esophageal cancer in Kenya, overall and for demographic/exposure groups. These fractions will also be compared to those for other risk factors, to feed into Kenya's next (due 2023) cancer control plan.

Planned Impact

We have provide a detailed description of the wider societal and academic impacts of our research in the Pathways to Impact document. We propose to quantify the role of exposure to household air pollution (HAP) from solid fuel use for cooking in the aetiology of esophageal cancer (EC) - a hitherto under-researched but significant risk factor (87% of households in Kenya rely on solid fuel for cooking) that particularly affects women and young children due to traditional domestic roles. This study will provide important findings in three main areas: (i) the evaluation of HAP as a risk factor for esophageal cancer (EC); (ii) primary prevention of EC in Kenya, through reduction of HAP exposures and (iii) if the magnitude of associations with EC found are strong, they can be incorporated into risk-prediction models for EC as a first step to identify individuals with high risk of EC in Kenya. The factors being studied are particularly attractive because modifications would have benefits for multiple health endpoints and they are relatively easy to communicate.
1. Examining HAP as an esophageal carcinogen which provide important results, not only for Kenya, but globally. Quantification of EC burden attributable to HAP exposure will allow prioritizing risk mitigation for EC.
2. Providing further evidence for the role of HAP in non-comminable disease burden supports global efforts to address the problem of HAP through supporting LMICs scale use of cleaner technologies and fuels. Our multidisciplinary team will produce robust scientific evidence, which will be carried forward for the development of effective policies and interventions to translate findings for the improvement of community health. HAP reduction is now on international and national health agendas, as endorsed by the World Health Assembly in 2015, and at the heart of Sustainable Development Goal 7 on Affordable and Clean Energy. The UK PI Dr Daniel Pope and investigator Dr Elisa Puzzolo are both highly engaged in efforts to promote global clean energy access for health, and are well placed to deliver the study's results to policy makers tackling HAP (CLEAN-Air(Africa) NIHR Global Health Research Group). Kenya has engaged with this agenda and situational analyses for transitioning to LPG as a clean fuel has been undertaken, engaging with multiple cross-sector stakeholders. The majority of the Kenyan population (87%) use biomass as their fuel source i.e. a population set to gain from a shift to LPG. With logistics and cost barriers, household use of LPG will be more challenging to achieve in the rural Kenyan setting in the short-term, but his population may benefit from improved biomass cook stoves in the interim. Further, in areas of very high EC risk, where extended families are themselves aware of multiple affected cases (and no yet-established genetic familial risk), individuals and local authorities may be incentivised to take action to reduce HAP exposures if HAP is a contributor to EC.
3. In addition to HAP, other EC risk factors assessed (e.g. alcohol, tobacco, oral health, hot beverage drinking) - are also either important risk factors for multiple NCDs, or are easily modifiable and communicated risk messages. The prospects for prevention are real for a cancer like EC. In Kenya, EC has been the 3rd most common cancer (both genders) for 40 years, with no sign of a decline in incidence. In 2012, there were an estimated 3400 new ECs diagnosed in Kenya, but with population ageing and expansion, by 2030, this annual patient burden will have more than doubled. If primary prevention through informed risk factor reduction could achieve a 25% reduction in incidence, by 2030 over 2000 new EC cases could be avoided each year. With low survival (median 5 months), this translates into approximately the same number of avoided deaths, about 15% of which occur at very young ages (<40 years).
 
Description We are currently in year one of the award and therefore do not have any key findings to report. We are however assisting our Kenyan research partner (Moi University) in disseminating findings from the first case control study conducted on oesophageal cancer in Eldoret - upon which HAP ESCCAPE has been built. With our partners from Moi University and the International Agency for Research in Cancer (IARC) we have extracted evidence of a suggestive hypothesis between household air pollution and oesophageal cancer in women and young men. Results indicate that self reported choice of fuel (and retrospective exposure to HAP from sleeping by the fire as a child) raise the risk of oesophageal cancer in these demographics. The findings are only suggestive at present and require confirmation through HAP ESCCAPE (our funded work). We will collaborate on published outputs from this dissemination (abstract and publication).

UPDATE (2021): We have been working closely with IARC to understand the potential mechanisms for the association between household air pollution (from solid fuel use) and esophageal cancer. This has involved detailed review of the literature on polycyclic aromatic hydrocarbons and cancer (upper airway and lung) and open discussion (e.g. journal clubs). We have also begun to explore the relationship between particle size (PM10 vs PM2.5) in the association through collaboration with partners from the US Research Triangle Institute (RTI). This work will be important in developing our analytical models of PM/PAH and case status (esophageal cancer) through the field work for HAP ESCCAPE. We are currently developing our analysis strategy.

The primary publication (described in the last report - 2020) is almost complete and ready for submission (Environmental Health Perspectives).

Our work on the impact of COVID-19 and associated control measures in Nairobi and Eldoret on household income and fuel insecurity (leading to greater use of polluting solid fuels and higher concentrations/ exposures of household air pollution) has now been published (Review Sustain Energy Rev 2021; Jul;144:None. doi: 10.1016/j.rser.2021.111018).

UPDATE (2022): First analysis of ESCCAPE data. As described in the 2021 update, a publication summarising the first 2 phases of the ESCCAPE study was drafted and submitted to Environmental Health Perspectives. The paper presented initial out analysis of the hospital based case-control study describing the association between exposure to smoke from combustion of wood (through cooking and heating at home) with the risk of esophageal cancer based on self-reported use of wood fuel and proxies for exposure (for example sleeping by an open fire when sleeping as a child and as an adult). These proxies for exposure were used to explore the hypothesis that chronic exposure to household air pollution (respirable particulate matter (PM2.5) and polycyclic hydrocarbons (PAHs)) was an important risk factor in the development of esophageal cancer in Kenyan adults; particularly in women who typically do not drink alcohol or smoke cigarettes (which are established risk factors for the disease). PAHs are chemicals known to have carcinogenic properties and inhalation of PAHs bound to PM2.5 facilitate transmission to the respiratory tract including the oesophagus. The analysis was supplemented with additional data collected on a urinary biomarker for an important PAH (Urinary 1-hydroxypyrene (1-OHP)) that is a reliable biomarker for total PAH exposure and was collected for ESCCAPE for a sample of control subjects. The data was analysed to establish whether levels of 1-OHP were higher in women than men, after adjustment, that would be consistent with the hypothesis that exposure to wood smoke was potentially an important risk factor for esophageal cancer in women. We found evidence of an increased association with wood smoke. The analysis found an association between various indicators of exposure to wood smoke and esophageal cancer including (i) use of wood as a primary fuel for cooking (OR=2.51; 95%CI = 1.61, 9.30), (ii) reported irritation of the eyes from smoke when cooking (OR=3.63; 95%CI = 1.61, 3.90) and (iii) sleeping by the open fire when as a child (OR=2.32; 95%CI=1.17, 4.62) and as an adult (OR=4.82; 95%CI =1.35, 17.2). Whilst adjustment was carried out for age to ascertain these exposures, further adjustment for covariates (including other risk factors such as alcohol and tobacco) led to non-statistically significant associations. We also found an association between sex and analysis of urinary PAH such that in control subjects the ratio of women to men for geometric mean 1- hydroxypyrene PAH concentrations was 2.27 (1.63, 3.17) - consistent with the hypothesis that wood smoke exposure was potentially a risk factor in this demographic. We received some detailed an informative feedback from peer reviews of the manuscript that have led to further analysis. Firstly we need to account for potential selection bias in relation to access to health care for Moi Teaching and Referral Hospital (the health facility for the project). Secondly it is very difficult to disentangle exposure to air pollution from use of wood for cooking (given that most of the participants (cases and controls) have been using wood all their lives and continue to do so - a very important observation. This is an issue because there is no clear reference group with which to compare the wood users (the reference being a composite of households currently using polluting fuels such as charcoal and kerosene and clean LPG for cooking). Any adjustment for cofounding and covariates beyond simple key confounders (e.g. age, assets) significantly reduces the power of effect estimates.

We have attempted to address these methodological issues through additional analysis bringing in new data from the HAP ESCCAPE recruitment that has boosted to the numbers of cases and controls to over 600 in each group. We are replicating the analysis (including multivariable analysis) to see if the associations hold. Unfortunately the issue of disentangling differential exposure remains difficult. The HAP ESCCAPE component was designed to take a life course of exposure to wood smoke to provide more variation in exposure to elucidate the relationship between increasing exposure and cancer risk. The life course analysis confirmed that participants have typically have been using wood for fuel for all their lives (recalling wood smoke exposure as children, whilst at school, during early adult life and currently - it has therefore not been possible to add additional granularity to the analysis. We have therefore engaged assistance from a statistician to create an exposure matrix which can create an exposure score based on self reported indicators of wood smoke found to be bivariately associated with cancer risk from initial modelling (exposure to more factors being associated with an increasing risk of cancer). The index is currently being analysed alongside the larger dataset for refinement of the manuscript.

One of the unique features of the HAP ESCCAPE project is the objective measurement of PM2.5 and PAHs undertaken in the field of the households of cases and controls. Whilst Covid-19 greatly impacted the collection of this data (as documented in the request for a no-cost extension) with a reduced sample size and data quality issues (damaged filters for collecting of particulates) we have now processed all the PM2.5 and PAH data by (i) fuel type and (ii) case-control status. This data will help us identify levels of a key range of health damaging PAHs (beyond 1-OHP) that can be attributed to combustion of wood fuel and translated into hazardous exposure. We have contracted US experts (Research Triangle Institute) to conduct analysis and dissemination of this data and to assist in the production of a peer reviewed journal article with us. To date there is no such analysis available in the literature.

Following publication of the article concerning energy security following Covid-19 in Eldoret and Nairobi (detailed in the 2021 update - Applied Energy), research that was conducted after field activities were ceased due to the pandemic, an additional research activity and analysis in Kenya was written up considering how the impacts on energy security from income insecurity during community lockdown from Covid-19 could be mitigated through through pay-as-you-cook smart meter technology to facilitate cleaner cooking. This research was conducted in an informal settlement in Nairobi that had contributed data used in the previous paper. The research showed that resource poor households that were able to cook cleanly with gas through pay-as-you-go technology continued to do so before, through and after community lockdown from the pandemic. This contrasted to households without the technology (a significant reduction in use of gas for clean cooking was observed with regression to use of charcoal and wood (polluting fuels) that could be purchased in small amounts. The results were published in Applied Energy 2021; 292: https://doi.org/10.1016/j.apenergy.2021.116769). The were timeliness of the findings and importance to the public health field concerning the health effects from exposure to household air pollution from reliance on solid fuels for cooking resulted in a special WHO feature with a case study on its website https://www.who.int/news-room/feature-stories/detail/making-clean-cooking-affordable-and-accessible-during-covid-19-pay-as-you-go-smart-meters-promote-health-equity-nairobi.

Whilst the main results from HAP ESCCAPE have been delayed through a variety of impacts from the Covid-19 pandemic, relating to field activities in Kenya and UK staffing, progress with dissemination is on track for dissemination in 2022 with outputs anticipated shortly after the 3 month no-cost extension period. Abstracts are being developed for the forthcoming 2022 ISEE conference (epidemiological hospital data on wood smoke and esophageal cancer risk) and ISES conference (patterns of PM2.5 borne PAH profiles from combustion of solid fuels in patients with and without esophageal cancer from the Kenyan ESCCAPE case-control study).

UPDATE March 2023: Results were presented at the ISEE 2023 Conference (Athens): https://ehp.niehs.nih.gov/doi/abs/10.1289/isee.2022.P-1072. These highlight the latest analysis indicative of an association between proxies for wood smoke exposure and increased risk of esophageal cancer. Our partner Dr Diana Menya (PI from Moi University) attended the international meeting of the ESCCAPE studies hosted by IARC in Tanzania (discussions are ongoing about the next steps for looking at use of wood for cooking/ heating and esophageal cancer). The meeting included all the ESCCAPE research countries (Malawi, Tanzania and Kenya) - agenda available: <2023_Moshi_agenda.pdf>
Exploitation Route HAP ESCCAPE - Kenya is very relevant to other ongoing IARC led case control studies in Tanzania and Malawi, that are looking at risk factors for oesophageal cancer in these countries and where there is a high proportion of the pollution relying on solid fuels for their household energy (the source of high exposures to household air pollution). The methods of enquiry to ascertain retrospective exposure to household air pollution detailed in our report are unique and designed to elicit comprehensive assessment of the risk of household air pollution for oesophageal cancer. These reports will be supplemented with objectively measured particulate matter (and extracted polyaromatic hydrocarbons) as a further validation. The data collection tools will be made available to partners from these other sites and, of course, can be used by IARC for future studies.

UPDATE 2021: We have begun discussions with senior colleagues at IARC about analytical modelling involving Population Attributable Fractions from household air pollution (and other key risk factors - e.g. hot beverages, smoking, alcohol). These are crucial for informing policy related prevention strategy.

UPDATE 2022: We have leveraged the funding through the partnership between Moi University and the University of Liverpool such that impactful activities beyond HAP ESCCAPE have been carried out relevant to the promotion of clean household energy to address non-communicable disease burden in Kenya. The relevance of the Covid-19 impact research in Eldoret and Nairobi has been described above. In addition Dr Menya (Moi University) and Prof Pope (University of Liverpool) conducted research in the study communities contributing to the HAP ESCCAPE project in Eldoret to investigate the impact of the Kenyan government's imposition of VAT on LPG to reclaim expenditure on the Covid-19 pandemic. The research demonstrated that many households were switching back to polluting solid fuels due to these associated price increases. Together with the Ministry of Health and Ministry of Energy we drafted a policy brief based on these results that is now being put before the Kenyan Treasury to reinstall zero-rating on LPG to help achieve the countries aspirational targets for clean cooking (transitioning away from polluting solid fuels) by 2030 under Sustainable Development Goal 7 "Universal Access to Clean Modern Energy".
Sectors Education,Energy,Environment,Healthcare
URL https://www.who.int/news-room/feature-stories/detail/making-clean-cooking-affordable-and-accessible-during-covid-19-pay-as-you-go-smart-meters-promote-health-equity-nairobi
 
Description Whilst the results of HAP ESCCAPE have not yer directly led to non-academic impact, findings from the analysis of the impacts of Covid-19 lockdown in terms of energy security and use of polluting fuels that were published in two academic journals have led to such impact. The research was carried out during cessation of research activities for HAP ESCCAPE partly in the study communities of Eldoret (the location of Moi Teaching and Referral Hospital and focus for the case-control study). Impacts from the research include (i) publication of an article in The Conversation that documented issues on income and energy security control measures for Covid-19 in Kenya and (ii) a case study developed by the World Health Organisation on the benefits of pay-as-you-go smart meter LPG for clean cooking in times of economic instability (now published on the WHO website).
First Year Of Impact 2021
Sector Energy,Environment
Impact Types Societal,Policy & public services
 
Description Community Health Worker Training on Air Pollution, Health and Prevention in Kenya
Geographic Reach National 
Policy Influence Type Influenced training of practitioners or researchers
Impact Whilst not directly linked to the HAP ESCCAPE project, our evaluation of CHV training in the new health module has identified two immediate case studies where households have transitioned from polluting solid fuels for cooking to clean LPG. This included a household using charcoal and experiencing symptoms of carbon monoxide poisoning and respiratory problems. The CHV responsible for this household explained the health risk from exposure to HAP (cancer, COPD, CO poisoning and cardiovascular disease) and then discussed prevention strategies - according the to training she had been given under the technical training module. The household immediately changed their cooking behaviour (cooking outside) before switching to an LPG cylinder for cooking. Her health symptoms immediately improved. She has delivered messaging within her community (an informal settlement in Uasin Gishu county near Eldoret) and 2 of her neighbours have now switched to clean cooking with LPG. Professor Pope visited the household and is drafting a case study that will form the basis of a video and publicity for CLEAN-Air(Africa).
 
Description Ministry of Health, Kenya
Geographic Reach Local/Municipal/Regional 
Policy Influence Type Contribution to a national consultation/review
Impact In its collaboration with the Kenyan Ministry of Health for CLEAN-Air(Africa) the University of Liverpool, Moi University and Amref International University have drafted a new training module for the Kenyan Community Health Workforce on Household air Pollution, health and prevention. This has been ratified by the Kenyan ministry of health and is currently being rolled out across the 47 counties of Kenya under Universal Health Coverage. In development of the new module (2019-2020) the Technical Working Group (Pope, Menya, Mwitari) provided a detailed account of important health outcomes relevant to the burden of disease in Kenya from household air pollution. Esophageal cancer and the HAP ESCCAPE project were discussed highlighting the priority of this outcome in Western Kenya. The results from HAP ESCCAPE will be presented to the technical working group when the study is complete. Early dissemination of results on household air pollution and health impacts are being presented from CLEAN-Air(Africa) in the study communities of Kesses and Turbo at a public and stakeholder engagement event (March 2020). The locations represent the wider community for the HAP ESCCAPE study with Moi Teaching and Referral Hospital being located centrally in Eldoret and the study will be presented to the community and health, clinical and administrative stakeholders at the event. The event will be reported on the CLEAN-Air(Africa) website www.cleanairafrica.com.
 
Description Online lecture
Geographic Reach Multiple continents/international 
Policy Influence Type Influenced training of practitioners or researchers
Impact Training is designed for African students, health practitioners, clinicians and wider academic and research groups to provide training in researching household air pollution and global health. The training includes specialistic lectures of influential epidemiological studies of key health outcomes and HAP ESCCAPE will form the topic one of these expert lectures. The training event will be launched in April 2021 from the CLEAN-Air(Africa) website. The expert lecture for HAP ESCCAPE describes use of a case control study design to investigate the role of household air pollution on esophageal cancer. This includes raising the profile of the relevance of the outcome in Western Kenya given the incidence and burden in the region and in particular the unknown aetiology in specific demographic groups (for example women and young children). Dr Iva Cukic will deliver the lecture on behalf of the HAP ESCCAPE team. The lecture will be placed alongside other seminal presentations of epidemiological study designs of core health outcomes associated with household air pollution.
URL https://www.cleanairafrica.com/
 
Title Objective Measurement of Household Air Pollution 
Description The original ESCCAPE case control studies estimated exposure to household air pollution (HAP) based on a relatively poor proxy of current fuel used for cooking and heating. HAP ESCCAPE seeks to objectively measure current exposure to the health damaging components of HAP that could affect respiratory health (particulate matter (PM) and have carcinogenic properties (polycyclic aromatic hydrocarbons (PAHs)). To do this we use state of the art monitoring equipment suitable for measuring HAP both in terms of concentration (e.g. indoors/ kitchen) and for personal exposures (PM2.5 - PM of 2.5 microns or less). UoL has a unique partnership with the Research Triangle Institute (RTI) who have developed small, lightweight gravimetric air pollution monitoring devices (MicroPEMs) that collect PM2.5 samples over specified recording periods (e.g. 24 hours) and measure real time levels of PM2.5 through laser light-scatter. The microPEMs that can measure average levels of PM2.5 over a 24hr period and, through careful management of the collected samples, allow extraction of levels of PM2.5-bound polycyclic aromatic hydrocarbons (PAH) for the sampling period. Analysis of PAHs (organic compounds that are causally implied in cancer aetiology) will help establish the mechanism of any association identified between HAP exposure and esophageal cancer. We have developed methods to deploy this instrumentation for minimal inconvenience for the study participants (microPEMs being warn on a harness or in an apron during the recording period). We have also developed standardised protocols for effective training of fieldstaff. For HAP ESCCAPE we aim to collect objective data on HAP from 50 cases and 50 controls that are either women or male younger than 40 years old (current recruitment is at 12 cases and 5 controls). Importantly, for some cases, it will no longer be possible to carry out the domestic tasks they previously undertook (including cooking) due to ill health. In this situation, domestic proxies (relatives or friends who have assumed the cooking responsibilities) will be recruited to wear the air pollution monitoring equipment for the recording period. Ultimately, we plan to work with IARC to (i) compare levels of HAP to WHO guideline levels and (ii) produce a current and lifetime HAP exposure prediction model which will then be applied to the full case-control study (625 cases and 625 controls). 
Type Of Material Improvements to research infrastructure 
Year Produced 2019 
Provided To Others? No  
Impact N/A 
 
Title Retrospective Exposure to Household Air Pollution Questionnaire 
Description In studies of household air pollution (HAP) and health, researchers have typically rely on a limited question set assessing current fuel and stove use as a poor proxy for exposure. Despite the potential for exposure misclassification, HAP has still been demonstrated to be associated with a wide range of health conditions, including cancer. For communities where the predominant current domestic fuel is biomass, use self-reported fuel/ stove use to infer exposure will make it difficult to tease out exposure-outcome relationships due lack of variability in levels of HAP exposure. Whilst this could be overcome by measuring actual levels of exposure objectively (as we are doing in a sub-sample for HAP ESCCAPE), such measurement is costly and intensive. Furthermore, in the aetiology of cancer we are interested in longer term exposure rather than current levels. Therefore, obtaining a differential and detailed life history of potential HAP exposure might be important in providing the granulation and chronicity of exposure necessary to understand these relationships. To attempt to address this we developed a comprehensive questionnaire designed to assess lifetime exposure to HAP, from cooking, heating and lighting sources. The first phase of questionnaire development involved consultation with leading international HAP experts (e.g. Dr Om Kurmi, McMaster University) following a structured literature review and thematic discussions within the energy, air pollution and health research group. Accordingly, we developed a pilot questionnaire, designed to link recollection of participants' exposure to potential sources of HAP with key life phases/ events (e.g. childhood, while in school, first job, etc.). The pilot questionnaire was administered to 30 individuals from our study community and was refined based on their feedback to develop the main study exposure questionnaire. This version was administered as a 'rolling pilot' in the study setting (hospital), to the first 20 cases and 20 controls, with minor amendments to the questionnaire being made based on their responses. We have now collected this retrospective exposure information from 100 cases and 60 controls out of the target 225 cases and 225 controls. The questionnaire is administered in a dynamic adaptive format - follow up questions are presented to the participant only when relevant based on their previous answers, to minimise the administration time. There are three broad components of the questionnaire. In the first component, the participant is asked about present cooking, heating and lighting devices, fuels and habits in their household, as well as whether they are the main cook in the household. If they confirm being the main cook, an additional set of questions is asked, assessing their cooking habits (e.g. how many times in a day/week they cook, what is the average cooking time per meal, etc). Next, the participant is asked about general HAP circumstances they can recall from their childhood (for example, whether they slept by the burning fire as a child in addition to cooking and heating habits if their childhood home). Finally, the questionnaire is designed to detect any significant changes in HAP circumstances between childhood and the present day, by asking a series of questions about any relevant changes and participant's age at change. This novel approach enables detection and subsequently modelling of each individual's specific circumstances in terms of HAP exposure during their lifetime. The questionnaire will be made publicly available for future use after the data collection phase is completed and metric characteristics of the questionnaire (reliability, validity) assessed. 
Type Of Material Improvements to research infrastructure 
Year Produced 2019 
Provided To Others? No  
Impact Unfortunately the life course analysis of differential exposure to household air pollution (wood smoke) afforded by the questionnaire has not helped in disentangling the extent of cumulative exposure in the cohort of cases and controls for HAP ESCCAPE with almost all homes currently cooking with wood (and indicating exposure through proximity to smoke as a child and adult) being exposed to wood smoke all their lives (from childhood, adolescence and adulthood). Without more variation over time (e.g. people changing their choice of cooking fuel to cleaner alternatives such as LPG) it is not possible to highlight more granularity on exposure to tease out exposure-response relationships. The tool is likely to be more effective in settings where there is a greater use of gas/ electricity for clean cooking (with earlier transition to cleaner modern energy) - e.g. urban settings. 
 
Title Additional Environmental Samples 
Description In addition to objective HAP measurements conducted on a subsample of 50 cases and 50 controls, we collect additional environmental samples in each household. The samples collected are: water used for drinking in the household; sample of two types of grains and beans used for cooking in the household; and a sample of soil from the garden used for growing vegetables or other crops. We are creating a biobank with support of Moi University environmental team, where these specimen will be preserved, and further funding will be sought to conduct relevant analysis in relation to other environmental factors implicated in esophageal cancer. 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? No  
Impact N/A 
 
Title Environmental Samples Biobank 
Description In addition to HAP samples (particulate matter (PM2.5) and polycyclic-aromatic hydrocarbons (PAH)) collected in the sub-sample of 50 cases and 50 controls for HAP ESCCAPE, we are collecting a number of environmental samples to test for potential carcinogens to explain the high occurrence of esophageal cancer in this setting. The samples include drinking water, two types of grains and beans used for cooking in the household and a sample of soil taken from the garden where vegetables or other crops are grown. Analysis of these samples is beyond the scope of the current study/ award but we are creating a biobank with support of Moi University environmental team and the British Geological Survey, where specimens will be preserved. We will seek additional funding to analyse these samples following completion of the HAP ESCCAPE study. 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? No  
Impact N/A 
 
Title Hospital Interviews Database 
Description The retrospective HAP questionnaire, described in detail above, is administered in the hospital setting as part of the comprehensive hour-long interview with all cases and controls recruited for the study, with details of demographics, medical history, and other esophageal risk factors assessed in addition to HAP. The questionnaire is administered by the interviewers in the hospital, using a tablet. Scanning a unique QR code generates a study ID number for each participant. The interviewers read out questions to the participants, and input their answers into the tablet. The database is automatically generated, and extracted as a .csv file for purposes of quality control and future analyses. The current database is stored on GDPR compliant IARC secure servers and will be merged with the original ESCCAPE case-control study (440 cases and 440 controls) for the purpose of harmonising variables on all risk-factors and increasing overall power of the analyses. 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? No  
Impact N/A 
 
Title Household Air Pollution Samples (PM2.5 and PAH) 
Description The RTI microPEM used in HAP ESCCAPE contain small pre-weighed filters that capture particulate matter (PM2.5) over a continuous 24h period (both concentrations (kitchens) and exposures (cases/ controls)). After use the filters are validated and cold stored before being cold shipped to our research partner in the US (Research Triangle Institute - RTI). The filter samples are then post-weighed to calculate the mass of PM2.5 deposited on the filter. During the final MicroPEM data processing the PM concentration derived from the filter, which is the gold-standard for exposure measurements, is used to post-correct the real-time light scattering nephelometer concentrations of the corresponding MicroPEM data files. The real-time patterns of exposure are summarized by calculating exposure cumulative percentile levels ranging from the 0.5 to 99.5 percentiles. In addition, the final data processing stop summarizes MicroPEM performance during each sample by calculating statistics for operating variables such as relative humidity, temperature, differential pressures, flow rates, etc. Samples with operating variables outside of the expected ranges are flagged for manual inspection for validity determination. Finally, each microPEM is fitted with an accelerometer to measure wearing compliance (for exposure measurement) and this metric is added to the complete dataset. RTI provide a full summary dataset for these values combining filter measurements with data downloaded directly from the microPEM. For HAP ESCCAPE this dataset will be supplemented with data pertaining to concentrations of and exposures to PM2.5-bound polycyclic aromatic hydrocarbons (PAHs). The PAH levels are calculated using a RTI-developed standard operating procedure and analysed for polycyclic aromatic hydrocarbons (PAHs) by gas chromatography-mass spectrometry (GC-MS). GC-MS results will be normalized to mass per unit sample volume (e.g. ng/mL) for direct comparison of PAH-exposure. 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? No  
Impact N/A 
 
Title Household Surveys Database 
Description The second phase of the study involves objective measurements of HAP in households of 50 selected cases and 50 matched controls. As described above, two short surveys are designed to provide context of the measurements: one administered prior to deploying HAP measuring equipment (pre-placement survey), and the other 24h later, after the sampling is completed and equipment removed (post-placement survey). The surveys are designed using an online survey platform Mobenzi, and are administered by the fieldworkers using an android mobile phone. Both questionnaires are read out loud to the participants (where applicable) or filled in by the field workers alone (when their perceptions are sufficient, e.g. recording the number of windows in the kitchen). The database is automatically generated within the Mobenzi platfom, and extracted as a .csv file for purposes of quality control and future analyses. 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? No  
Impact N/A 
 
Title Objective HAP measurements 
Description As described above, two MicroPEM devices are deployed in each household: one for measuring kitchen concentrations, and the other measuring personal exposure to PM2.5. They contain a small pre-weighed filter that captures the HAP particles over a continuous 24h period. All filter samples will be post-weighed to calculate the mass of particulate matter deposited on the filter. During the final MicroPEM data processing the PM concentration derived from the filter, which is the gold-standard for exposure measurements, is used to post-correct the real-time light scattering nephelometer concentrations of the corresponding MicroPEM data files. The real-time patterns of exposure are summarized by calculating exposure cumulative percentile levels ranging from the 0.5 to 99.5 percentiles. In addition, the final data processing stop summarizes MicroPEM performance during each sample by calculating statistics for operating variables such as relative humidity, temperature, differential pressures, flow rates, etc. Samples with operating variables outside of the expected ranges are flagged for manual inspection for validity determination. In addition to these measures of concentration and exposure to PM2.5, an additional analysis of PM2.5bound polycyclic aromatic hydrocarbons (PAHs) will be conducted. The PAH levels are calculated using a RTI-developed SOP and analysed for polycyclic aromatic hydrocarbons (PAHs) by gas chromatography-mass spectrometry (GC-MS). GC-MS results will be normalized to mass per unit sample volume (e.g. ng/mL) for direct comparison of PAH-exposure. 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? No  
Impact N/A 
 
Description International Agency for Research on Cancer (IARC) 
Organisation International Agency for Research on Cancer (IARC)
Country France 
Sector Academic/University 
PI Contribution We are collaborating closely with the International Agency for Research on Cancer (IARC) who have coordinated three existing case-control studies of esophageal cancer in East Africa in Kenya, Tanzania and Malawi - collectively called the ESCCAPE studies. Our study was designed to address a gap in the evidence identified from these studies - namely the role of exposure to household air pollution (HAP) in the aetiology of esophageal cancer (HAP ESCCAPE), building on the original ESCCAPE study in Kenya. The ESCCAPE studies all focussed on individual and environmental determinants of esophageal cancer (e.g. smoking, alcohol, contaminated food/ water), but for a particular demographic (women, young men) these 'traditional' risk factors appeared to have less relevance. In all settings a substantial number of the population rely on traditional use of solid fuels for cooking and heating their homes and the household air pollution (HAP) arising from combustion of these fuels was a postulated risk factor in the aetiology of the disease, particularly in this demographic. The University of Liverpool (UoL) joined Moi University (the lead organisation for ESCCAPE Kenya) to extend the case control study to elucidate this relationship and accordingly partnered with IARC. The UoL team brings significant expertise and years of experience in HAP, health and prevention research. This has a two-fold significance: Firstly, helping to formulate research hypotheses and design the most appropriate statistical models using existing data from the ESCCAPE studies, as well as the analytical skills to conduct such analyses. Secondly, designing the tools, methods of enquiry and objective assessment of HAP to accurately measure exposure to health damaging components of air pollution including its carcinogenic properties. In designing the analytic methods for understanding exposure-outcome relationships, the collaboration between lead members of the UoL and IARC teams has been invaluable and has included a one-week analysis meeting at IARC Headquarters in Lyon. Prior to the meeting the head of the UoL team gave a departmental seminar to the IARC team describing the role of HAP in human cancer and wider work by the UoL research theme. UoL is currently providing statistical expertise to IARC and Moi teams in interrogating existing datasets to tease out retrospective exposure to HAP as a predictor of oesophageal cancer, adjusting for other covariates. Moving forward, UoL will collaborate with IARC to coordinate relevant dissemination from the three ESCCAPE studies in relation to HAP and oesophageal cancer across East African settings.
Collaborator Contribution IARC is the leading international institution for conducting research on cancers and collating present evidence on behavioural and environmental risk factors. They have an extensive team of cancer epidemiologists, geneticists and environmental health experts and provide unique knowledge and expertise in many of the most prevalent cancer conditions in Sub Saharan Africa. After helping elucidate the significant prevalence of esophageal cancer in Western Kenya, IARC collaborated with our partner, Moi University, on the first ESCCAPE study of esophageal cancer in Western Kenya upon which the present study is built. At the same time two other ESCCAPE studies were launched, again coordinated by IARC, that were conducted in contexts with high prevalence rates of esophageal cancer (settings in Tanzania and Malawi). Through IARC, we have begun networking with the academic institutions leading these other case-control studies and to discuss the role of HAP as a risk factor more widely across settings. Within IARC, the ESCCAPE research is being coordinated by the Section of Environment and Radiation whose remit is to investigate environmental, lifestyle, occupational, and radiation-related causes of cancer in human populations. During the course of HAP ESCCAPE we have established a close working relationship with its Section Head (Dr Joachim Shuz) and Deputy Section Head (Dr Valerie McCormack) who have provided substantial input into the study design of HAP ESCCAPE and helped discuss/ address challenges of fieldwork in relation to measurement of HAP in the Kenyan context. In October 2019 IARC hosted the HAP ESCCAPE project coordinator (Dr Iva Cukic) for a week, to review existing ESCCAPE datasets and plan and conduct analyses for a joint publication (in progress) based on the original Kenyan case-control study. The analyses provided important, suggestive evidence for the role of HAP in increasing the risk of esophageal cancer paving the way for much needed quantitative evidence (as provided by HAP ESCCAPE). An abstract on this analysis has been submitted by UoL/ IARC/ Moi for the International Society for Environmental Epidemiology (ISEE) Conference in Washington, August 2020 - a key conference for other research activities conducted by the UoL team. In February 2020, IARC organised and hosted a two-day meeting of the ESCCAPE studies in Blantyre, Malawi. Attendees included principle investigators and representatives of all ESCCAPE participating sites, in Kenya, Malawi and Tanzania. Progress of each of the studies was reported, and challenges and experiences exchanged along with some initial results. UoL (Iva Cukic) presented on the current progress and challenges of HAP ESCCAPE. The event helped developed further collaboration between UoL and the ESCCAPE teams in Malawi and Tanzania, and following the meeting UoL are leading plans to harmonise available HAP variables for a joint analysis of HAP indicators in relation to esophageal cancer diagnosis for a multi-site study.
Impact Residential strategic analysis meeting hosted at IARC HeadQuarters, Lyon (October 2019) - presentation by UoL theme lead (Professor Daniel Pope). Representation by HAP ESCCAPE project coordinator (Dr Iva Cukic). International steering meeting for the ESCCAPE studies (hosted by IARC - Blantyre, Malawi (February 2020)) - presentation by HAP ESCCAPE project coordinator with strategic discussions on cross-centre analysis of solid fuel use (wood and charcoal) as a consistent risk factor for women and younger people for esophageal cancer in Malawi, Tanzania and Kenya. Presentation of the HAP ESCCAPE study and wider ESCCAPE collaboration at the International Society for Environmental Epidemiology (ISEE) 2020 annual conference. Follow-up with lead researchers in the field of air pollution to discuss logistics of retrospective cumulative exposure to household air pollution as a risk factor for esophageal cancer (presentation March 2020). The HAP ESCCAPE coordinator has led development of a publication based on the original Kenya ESCCAPE study (first case control study) that provided the background to HAP ESCCAPE. This paper is the results of a collaboration between Moi University, IARC and the University of Liverpool (final stages before submission).
Start Year 2019
 
Description NIHR CLEAN-Air(Africa) Global Health Research Unit - Co-Directorship 
Organisation Kenyan Institute for Medical Research (KEMRI)
Country Kenya 
Sector Public 
PI Contribution I am Director of the NIHR CLEAN-Air(Africa) Global Health Research Group and have developed a partnership with Dr James Mwitari of the Centre for Respiratory Disease Research at KEMRI, Nairobi. Dr Mwitari is co-Director of our future CLEAN-Air(Africa) Unit that will be led by the University of Liverpool as the host institution.
Collaborator Contribution Through KEMRI, that is a government organisation, we are engaging closely with the Ministries of Energy and Health to steer the clean cooking agenda in Kenya. We will be establishing an Air Pollution Centre of Excellence at KEMRI for all research related activities on the health effects of ambient and household air pollution. Dr Mwitari will steer these activities. This includes hosting of the MoH Technical Working Group on Climate, Health and Energy -it is anticipated that the group will facilitate all research activities in the country on clean household energy and health.
Impact The collaboration is multidisciplinary (mixed methods implementation research and evidence to policy; anthropology; epidemiology; public health; behavioural sciences).
Start Year 2020
 
Description RTI 
Organisation RTI International
Country United States 
Sector Charity/Non Profit 
PI Contribution UoL co-ordinated a training for the local team at MOI University, Eldoret, with the aim of both training the field workers in household air pollution measurements for the purposes of the HAP ESCCAPE project, and more broadly, building capacity for household air pollution measurements, thus enabling the local team to continue conducting crucial research in HAP far beyond the scope of the current study. The week-long training included a comprehensive theoretical and practical component focusing in Stove Use Monitoring, HAP measurements (PM 2.5 concentrations/ exposures) using microPEM and LASCAR devices, as well as emissions work for the field team. This comprehensive training in a range of methodologies and equipment The field team is further trained in administering mobile surveys via the Mobenzi platform, to accompany household measurements with supplementary survey data
Collaborator Contribution RTI are US based world leading experts in the HAP measurement technology and research. They contributed with their HAP measurement technology (microPEM and LASCAR devices), and provided expert training for the field workers in Eldoret. Furthermore, as data collection commenced RTI conducts weekly quality control check of the collected field data. Finally, RTI continues to actively engage in the discussions with the UoL team about the study design, hypothesis generation, and further planning on the topics of air pollution measurements, both particular matter and PAH chemical compound to assess possible biological mechanisms at play.
Impact Formal training of the field workers and the Eldoret academic and research team in the topics of household air pollution field measurements - as described in other sections of the form.
Start Year 2019
 
Description Community Health Worker Training Initiative - Air pollution, health and prevention 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Other audiences
Results and Impact The Universities of Liverpool and Moi (HAP ESCCAPE partners) have co-developed a national training initiative for community health workers in Kenya through the CLEAN-Air(Africa) Global Health Research Group and engagement with the Ministry of Health. In November 2021 and February 2022 workshops were held to train community health volunteers in the health effects of household air pollution and prevention strategies, in preparation for the launch of population awareness raising and national prevention through rollout of training across Kenya's 47 counties. Evidence from the ESCCAPE studies were included as evidence of how household air pollution might be associated with key cancers, beyond lung cancer where a casual relationship is established. Understanding the relationship of household air pollution with cardiovascular and respiratory disease is crucial for community health risks to communities so research evidence is important to substantiate training. Prof Pope coordinated both training events with supervision from Dr Menya. The rollout of the national training initiative is ongoing and being steered by the University of Liverpool, KEMRI and Moi University.
Year(s) Of Engagement Activity 2021,2022
 
Description Public and stakeholder engagement event 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Public/other audiences
Results and Impact A public and stakeholder engagement event will be held in March 2021 in Kesses and Turbo (study communities for HAP ESCCAPE and CLEAN-Air(Africa). Members of the community from each sub-county, academics, clinicians (including from Moi Teaching and Referral Hospital) and senior health and administrative officials will be in attendance (as well as having ministry of health representation). The event is centred on findings in relation to household air pollution and health and including presentation of the HAP ESCCAPE study.
Year(s) Of Engagement Activity 2021