You are here

  • Home
  • Archive
  • Volume 77, Issue 11
  • Multidimensional energy poverty and acute respiratory infection in children under 5 years of age: evidence from 22 low-income and middle-income countries

Article Text

Article menu

Download PDFPDF

Original research
Multidimensional energy poverty and acute respiratory infection in children under 5 years of age: evidence from 22 low-income and middle-income countries
  1. Merel Stevens1,
  2. Junwen Yang-Huang1,
  3. Daan Nieboer1,
  4. Shuang Zhou2,
  5. Kwabena Agyapong Osman3,
  6. Hein Raat1,
  7. Amy van Grieken1
  1. 1 Public Health, Erasmus Medical Center, Rotterdam, The Netherlands
  2. 2 Department of Maternal and Child Health, Peking University, Beijing, China
  3. 3 Departement of Child Health, University of Ghana Medical School, Accra, Ghana
  1. Correspondence to Dr Amy van Grieken, Public Health, Erasmus Medical Center, Rotterdam 3015GD, The Netherlands; a.vangrieken{at}erasmusmc.nl

Abstract

Background In low-income and middle-income countries (LMICs), energy poverty has predominantly been studied from the unidimensional perspective of indoor air pollution. Acute respiratory infection (ARI) in children under 5 years of age is the most important disease associated with indoor air pollution attributable to solid fuel use in LMICs. This study aimed to extend the existing knowledge on the association between energy poverty and ARI among children under 5 years of age in LMICs, by adopting a multidimensional perspective.

Methods Using Demographic and Health Surveys from 22 LMICs, data from 483 088 children were analysed (mean age 2.00 years (SD 1.41); 51.3% male). Energy poverty was measured using the Multidimensional Energy Poverty Index (MEPI) (range 0–1), which comprises five dimensions of essential energy services. Binary logistic regression models were conducted to study the association between MEPI and ARI, adjusting for child, maternal, household and environmental characteristics.

Results A 0.1 increase in MEPI score was associated with greater odds of ARI (aOR 1.05; 95% CI 1.04 to 1.07). Likewise, MEPI indicators using biomass for cooking (aOR 1.15; 95% CI 1.07 to 1.23) and lack of access to electricity (aOR 1.17; 95% CI 1.10 to 1.26), entertainment/education appliances (aOR 1.07; 95% CI 1.02 to 1.13) and household appliances (aOR 1.12; 95% CI 1.04 to 1.21) were associated with greater odds of ARI.

Conclusion Multidimensional energy poverty was associated with greater odds of ARI in children under 5 years of age living in 22 LMICs. Hence, our findings justify the design and implementation of interventions that address energy poverty from a multidimensional perspective, integrating energy affordability and accessibility.

  • poverty
  • health surveys
  • respiratory tract infections
  • child health

Data availability statement

Data are available in a public, open access repository. All data relevant to the study are included in the article or uploaded as online supplemental information. All data relevant to the study are included in the article or available as online supplemental material. The data used in the analyses are publicly available, anonymised and geographically scrambled to ensure confidentiality. More information on DHS can be found at 'https://dhsprogram.com/ (accessed on 8 August 2022)', where survey datasets can be obtained.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/jech-2023-220540

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Regardless of growing interest and awareness of the impact of energy poverty on children’s respiratory health, studies on this association in low-income and middle-income countries (LMICs) have mainly relied on unidimensional measures of energy poverty, with a focus on indoor air pollution. This study extends the existing knowledge by adopting a multidimensional perspective on energy poverty, which integrates aspects of energy accessibility and affordability, and its association with acute respiratory infection.

    WHAT THIS STUDY ADDS

    • This study demonstrates that multidimensional energy poverty is associated with acute respiratory infection in children under 5 years of age in LMICs.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • The findings encourage the design of interventions that improve both energy affordability and accessibility. Besides the implementation of clean cooking practices, access to electricity and ownership of household, education and entertainment appliances could equally contribute to the mitigation of energy poverty and its effects on children’s respiratory health in LMICs.

    Background

    Energy poverty has been associated with a diverse range of adverse health impacts, including respiratory disease in children.1 2 A recent definition that aims to overcome geographical barriers, considers energy poverty as the inability of a household to secure a socially and materially required level of energy services in the home.2 Globally, a consensus has been reached that reducing energy poverty is fundamental to socioeconomic development and health.3

    Both accessibility and affordability are relevant to the concept of energy poverty. In high-income nations, energy poverty is primarily addressed from an affordability perspective, based on household energy expenditures. In low-income and middle-income countries (LMICs), energy poverty is mainly considered as a lack of accessibility to modern forms of energy, such as clean cooking fuels.4 5 Indoor air pollution in LMICs is largely caused by solid fuel combustion for heating and cooking.6 Worldwide, around 2.4 billion people use solid fuels for cooking, most of them residing in LMICs.7 The majority of studies in LMICs address energy poverty from the unidimensional perspective of indoor air pollution.8 A more comprehensive way to capture energy poverty is essential for studies in LMICs since households might have access to modern energy services, yet they are unable to afford adequate levels of energy services.

    Acute respiratory infection (ARI) is an infection of the upper or lower respiratory tract, characterised by symptoms such as cough and shortness of breath.9 ARI in children under 5 years of age is the most important disease associated with indoor air pollution attributable to solid fuel use, with the highest burden occurring in LMICs.10–12 In fact, almost half of all deaths due to ARI among children under 5 years of age is caused by indoor air pollution.6 ARI is also associated with family socioeconomic status, an important predictor of energy affordability.13 14 A multidimensional measurement of energy poverty comprehends a more accurate picture of the extent and severity of the inequality in ARI experienced by children in LMICs.

    The Multidimensional Energy Poverty Index (MEPI), established by Nussbaumer et al 5 offers a wider framework to study the effects of energy poverty on ARI in children. It uses a multidimensional approach to capture a household’s deprivation in modern energy access, linked with the intensity and incidence of energy poverty. It, therefore, integrates aspects of both energy accessibility and affordability, by complementing commonly used indicators of electricity access and cooking fuel use with telecommunication means and household appliance ownership.5

    This study aims to extend the existing knowledge on the impact of energy poverty on ARI among children under 5 years of age in LMICs, by adopting a multidimensional perspective on energy poverty. Data from children under 5 years of age who responded to the Demographic and Health Survey (DHS) within the past 10 years are used to assess the association between multidimensional energy poverty and ARI.15 In addition, the association of the individual MEPI indicators with ARI is assessed.

    Methods

    This study followed the Strenghtening the reporting of observational studies in epidemiology (STROBE) cross-sectional reporting guidelines (online supplemental material 1).16

    Supplemental material

    Study design

    A community-based cross-sectional study was conducted across 22 LMICs. Data from the most recently available population-based DHS within the past 10 years (2012–2022) were used.15 DHS collects data from nationally representative, randomly selected samples of women of reproductive age (15–49) and their children (0–5 years) using the household and woman’s questionnaires. Each country follows a two-stage stratified cluster sampling technique, with probability proportional to the cluster size. Households are then randomly selected in each cluster. The surveys have large sample sizes (5000–30 000 households) and are conducted every 5 years. Standardised questionnaires are used with fieldwork supported by the United States Agency for International Development. Each survey round consists of a household questionnaire used to collect information on characteristics of the household’s dwelling unit and of usual residents and visitors. It is also used to identify eligible members of the household for the woman’s questionnaire. Data from the household and woman’s questionnaires were merged using Integrated Public Use Microdata Series DHS to obtain a working dataset for this study.17

    Study population

    The first selection of DHS surveys was made based on (1) DHS survey data available from within the previous 10 years (2012–2022), (2) children living in LMIC according to the World Bank classification,18 (3) the presence of MEPI indicators [5], (4) presence of the variables cough and shortness of breath, used to compose the outcome variable ARI according to the definition adopted by the UNICEF Multiple Indicator Cluster Surveys19 and (5) presence of relevant explanatory variables. The first selection included 410 567 households with 633 542 children from 31 LMICs.

    To obtain the final sample, visitors (n=20 953), countries (n=9) and children (n=33 773) with missing (in)dependent variables were excluded. The final sample included 483 088 children from 22 countries, living in 314 002 different households (online supplemental material 2).

    Supplemental material

    Acute respiratory infection

    The outcome of interest is the presence of ARI among all children under 5 years of age. ARI (yes/no) is defined as a cough accompanied by short, rapid breathing, according to the UNICEF Multiple Indicator Cluster Surveys, in the 2 weeks preceding the survey.19 Presence of cough and shortness of breath were therefore combined to compose the outcome variable ARI. The measure for ARI was modelled as a binary outcome (yes/no).

    To identify the presence of a cough, mothers were asked the following question: ‘Has (name) had an illness with a cough at any time in the last 2 weeks?’ Answer options were ‘yes’, ‘no’ and ‘don’t know’. Participants who answered ‘don’t know’ were excluded from the analysis.

    From 2012 to 2014, mothers of children who had a cough in the past 2 weeks were asked ‘when (name) had an illness with a cough, did he/she breathe faster than usual with short, rapid breaths or have difficulty breathing?’. From 2015 onwards all mothers of children, regardless of whether they experienced a cough in the last 2 weeks, were asked ‘has (name) had fast, short, rapid breaths or difficulty breathing at any time in the last 2 weeks?’. Answer options in both cases were ‘yes’, ‘no’ and ‘don’t know’.

    Cases that did not report the presence of a cough in the past 2 weeks with or without shortness of breath were categorised as not having ARI. Cases that had a cough but answered ‘don’t know’ to the shortness of breath question were excluded from the analysis.

    Multidimensional energy poverty

    MEPI was adopted to measure multidimensional energy poverty, which consists of five dimensions of essential energy services and includes six indicators: using biomass for cooking, indoor pollution, electricity access, household appliance ownership, entertainment/education appliance ownership and telecommunication means (table 1). Whereas the indicator ‘using biomass for cooking’ determines the release of polluting substances while cooking, indoor pollution further specifies the level of exposure to polluting substances inside the dwelling, which strongly increases when cooking inside in the absence of a separate kitchen.5 A modified MEPI was used due to missing values in the indicator ‘household has a separate room used as kitchen’ (48.1%) and the indicator ‘household has a television’ (46.3%). Cases with relevant data for at least one of the two variables were included in the analyses (table 1). Indoor pollution was determined by the single variable ‘where cooking is usually done (inside/outside)’ when ‘household has a separate room used as kitchen’ was missing. Entertainment/education was determined by the single variable ‘household has a radio (yes/no)’ when ‘household has a television’ was missing.

    View this table:
    Table 1

    Dimension, indicator and weighting of the Multidimensional Energy Poverty Index (MEPI)

    A household was classified as energy poor if the sum of assigned weights to each dimension exceeds a 0.30 cut-off value as defined by Nussbaumer et al 5

    Covariates

    Covariates were selected based on previous literature and availability in the DHS dataset.20 21 Child’s characteristics included age, sex (male, female), birth order (first-born, not first-born), breastfeeding status (ever breastfed/still breastfeeding, never), vitamin A supplementation in the last 6 months (yes, no), BCG vaccination (yes, no), low birth weight (yes, no) and treatment sought for fever/cough (yes, no). Maternal characteristics included age and education level (no education, primary, secondary, higher). Household characteristics included household smoking (never, daily, more than daily), source of drinking water (improved, unimproved), shelter index (low (0–1), intermediate (2–4), high (5–6)) and crowding (yes, no). Following previous work, the shelter index represents the sum of floor, wall and roof material, where natural material was coded as 0, rudimentary as 1 and finished as 2.22 Crowding was defined as more than 3 persons per room used for sleeping, a modified definition by UN Habitat.23

    Environmental characteristics included urban–rural status (urban, rural) and rainy season at the time of the interview (yes, no). The variable rainy season was included as a contextual factor to indicate whether the interview took place during the rainy or dry season, with information obtained from the World Bank Climate Change Knowledge Portal.24 Country of residence was considered as a covariate to compensate for differences across countries.

    Statistical analysis

    Descriptive statistics were used to characterise the sample, and summarised by the number of cases (n) and percentages (%) for categorical variables and mean and SD for continuous variables. The association between ARI as the main outcome variable and MEPI/energy poverty status as the main determinant was assessed using binary logistic regression. A unit of change of 0.1 for MEPI was used in the regression model. The regression model was adjusted for child’s age, child’s sex, first-born, vitamin A supplementation, BCG vaccination, treatment sought for fever/cough, low birth weight, breastfeeding status, maternal age, maternal education, shelter index, drinking water source, household smoking, crowding, urban–rural status, rainy season and country. Model fit was examined using Nagelkerke’s R2. Tests for multicollinearity were performed using variance inflation factors.

    The interaction effect between MEPI and child’s age, child’s sex, low birth weight, first-born, shelter index, household smoking, crowding, urban–rural status, rain season and country was evaluated by adding interaction terms to the adjusted models.10 14 25–28 The two-sided significance threshold, after Bonferroni correction for multiple testing, was set at 0.05 (original p value)/10 (number of interaction tests performed)=p=0.005. Stratified logistic regression analyses were run for the variables with statistically significant interaction terms. Bonferroni correction was applied for multiple testing per subgroup (0.05 (original p value)/8 (number of subgroups)=p=0.006).

    Eight of the 17 covariates had missing values for ≥2% of the sample: child’s age (5.5%), vitamin A supplementation (7.1%), breastfeeding status (13.4%), BCG vaccination (16.4%), treatment sought for fever/cough (77.7%), low birth weight (41.0%), shelter index (4.4%) and household smoking (2.1%). Multiple imputation (n=10 imputations) was used to impute the missing data using the chained equations method in R V.4.2.0.29 Rubin rules were used to obtain pooled estimates of ORs and their 95% CIs. Statistical significance in the adjusted model was set at p<0.05.

    All statistical analyses were performed using IBM SPSS Statistics for Windows, V.28.30 The regression analyses accounted for weighting, clustering and stratification using the Complex Samples Package in SPSS. Sampling weights were applied to the regression model to adjust for disproportionate sampling and non-response. The analyses considered the hierarchical structure of the sample. The clustering effect was accounted for by using the unique sample-case primary sampling unit identifier variable. The stratified sampling design was accounted for by using the unique cross-sample sampling strata variable.17

    Results

    Population characteristics

    The mean age of children was 2.00 years (SD 1.41); 51.3% were male; 25.7% resided in urban areas and 54.6% of the children lived in the South Asia region at the time of the survey, followed by 19.4% in Eastern Africa and 16.9% in Western Africa. In total, 38.1% of mothers had no education and 23.8% had a primary education level (online supplemental table 1). The mean MEPI was 0.51 (SD 0.23); 83.2% of children were considered to live in energy poverty conditions (table 2). In the overall sample, 7.8% of children suffered from ARI during the 2 weeks preceding the survey (online supplemental table 1).

    Supplemental material

    View this table:
    Table 2

    Energy poverty according to MEPI

    All variance inflation factors were lower than 4, suggesting low multicollinearity. No violation of basic assumptions for regression was found.

    Association between MEPI and ARI

    The adjusted model indicated that every 0.1 increase in MEPI was associated with greater odds of ARI in children under 5 years of age (aOR 1.05; 95% CI 1.04 to 1.07; R2 0.08) (table 3). Similarly, living in energy poverty was positively associated with ARI (aOR 1.18; 95% CI 1.09 to 1.27; R2 0.08).

    View this table:
    Table 3

    Associations between MEPI and ARI

    In addition, the association between the individual MEPI indicators and ARI was studied (table 3). The adjusted model demonstrated a positive association between using biomass for cooking and ARI (aOR 1.15; 95% CI 1.07 to 1.23; R2 0.08), between lack of access to electricity and ARI (aOR 1.17; 95% CI 1.10 to 1.26; R2 0.08), between lack of entertainment/education appliances and ARI (aOR 1.07; 95% CI 1.02 to 1.13; R2 0.08) and between lack of household appliance ownership and ARI (aOR 1.12; 95% CI 1.04 to 1.21; R2 0.08).

    Interaction analyses

    A significant interaction term was found for ARI between MEPI and child’s age (p<0.001) and MEPI and country (p<0.001). Stratified analyses (table 4) indicated a significant aOR for the association between MEPI and ARI among children aged 0–2 years (aOR 1.07; 95% CI 1.05 to 1.09) and 2–5 years (aOR 1.04; 95% CI 1.03 to 1.06). Stratified analyses for region showed a significant association between MEPI and ARI in South Asia (aOR 1.05; 95% CI 1.03 to 1.07), Eastern Africa (aOR 1.07; 95% CI 1.05 to 1.09) and Central Africa (aOR 1.10; 95% CI 1.04 to 1.16).

    View this table:
    Table 4

    Associations between MEPI and ARI by child age and region

    Discussion

    Our large cross-sectional, multicountry study showed a positive association between multidimensional energy poverty and ARI among children under 5 years of age. Our study also demonstrated that biomass cooking fuels, lack of access to electricity and lack of entertainment, education and household appliances are associated with increased odds of ARI. The results of this study reaffirm the critical role of improving the accessibility and affordability of modern energy services for households to address the burden of respiratory diseases among children in LMICs.

    Our results are consistent with previous studies that used other proxies for households’ access to modern energy services, such as electricity access, and its effect on health outcomes of children.31 32 In addition, our study extends the evidence stemming from a comparable study based solely on DHS data from Uganda, by confirming the association between multidimensional energy poverty and ARI across 22 LMICs.

    A large body of research confirms the association, observed in this study, between the MEPI indicator using biomass for cooking and ARI in children under 5 years of age.26 33 34 Likewise, a number of studies have reported a positive association between indoor air pollution and ARI among children.11 33 Contrary to expectations, this association was not evident in our study. A modified indicator was used for indoor air pollution to deal with missing values, which may affect its accuracy. Further research into the role of indoor air pollution to capture multidimensional energy poverty is warranted.

    The association between the MEPI indicator electricity access and ARI, observed in this study, is also confirmed by a recent review on the link between electricity and health in LMICs. Besides lowering the need for polluting fuels, access to electricity is considered to affect respiratory disease through increased wealth, improved sanitation and hygiene practices and increased exposure to health information.32 The negative association between the MEPI indicators entertainment/education and household appliance ownership and ARI in this study confirms the importance of overall improvement in family wealth. To illustrate, television and radio ownership improve exposure to environmental and health information for caregivers and their children.35 The MEPI indicator telecommunication ownership could be associated with ARI through similar patterns. Further research is needed to confirm this theory.

    The key contribution of this study is the use of the MEPI as a proxy for household deprivation in modern energy access, and its association with ARI among children across 22 LMICs. This study, therefore, adds to the existing knowledge on the association between energy poverty and ARI in children in LMICs, by adopting a multidimensional perspective that integrates energy affordability and accessibility. Not only energy accessibility (biomass cooking fuels, access to electricity), but also energy affordability (lack of entertainment/education and household appliances) is adversely associated with ARI. MEPI overlaps partially with the DHS Wealth Index, that is, constructed using a similar, but wider set of household assets and services.15 MEPI represents a more specific set of indicators that allows to accurately measure and study energy poverty as a distinct form of poverty. This is of importance, since energy is a basic necessity for meeting fundamental needs such as cooking, heating and access to clean water. Lack of access to modern energy services hinders human development, affecting health, education and overall quality of life. In addition, energy poverty is often linked to broader socioeconomic inequalities. Certain marginalised populations may face barriers to accessing affordable and reliable energy services, which exacerbates existing inequalities. Moreover, energy poverty is closely tied to environmental sustainability. The reliance on polluting energy sources contributes to deforestation and climate change. A transition towards sustainable energy sources is therefore urgently needed.36 A distinct index to study energy poverty helps to better understand its unique challenges and develop inclusive policies targeted at reducing energy poverty and its impact on health and well-being. The results of this study in particular encourage the design of interventions that improve (equity in) both energy affordability and accessibility. Current interventions to reduce energy poverty and its health effects in LMICs are mainly targeted at the transition to clean cooking fuels and stoves.37 Besides the implementation of clean cooking practices, access to electricity and ownership of entertainment, education and household appliances could equally contribute to the mitigation of energy poverty and its effects on children’s respiratory health in LMICs.

    This study has several strengths. First and foremost, MEPI is an innovative and robust tool that was constructed according to extensive and detailed microdata stemming from DHS household surveys.5 The dataset used in this study contained a large sample size (n=483 088) and enabled for comparison across 22 countries. Furthermore, confounders and competing risk factors were comprehensively addressed by drawing on a published causal diagram of ARI determinants.20

    The findings of this study must be considered in the context of some limitations. The outcome of this study, ARI, was self-reported by mothers over a 2-week recall period instead. Although the consistent use of maternal-reported ARI across all surveys allowed for comparison, misclassification due to recall bias or inadvertent maternal awareness might be present. Second, the cross-sectional nature of this study did not allow for inferring causal relationships. Furthermore, studies with very large sample sizes tend to generate statistically significant results more easily, in the absence of any practical significance.38 Bonferroni correction was applied in this study to reduce the chance of obtaining false-positive results. Nonetheless, future studies with appropriate power and corresponding sample size might be warranted. The evaluation of model fit through Nagelkerke’s R2 suggests that the independent variables in our model have a relatively limited ability to explain the variance observed in ARI. Although relevant to consider when interpreting our results, it is important to realise that ARI is a multifaceted public health issue that is difficult to capture in a single model.39 Although the overall model fit may be poor, individual predictors may still hold considerable significance when investigating such complex and multifactorial phenomena. Finally, surveys from the past 10 years were included in this study. The current era demands for transition to clean and sustainable energy sources. The containment of surveys to the past decade ensures the adequacy of our results and its policy implications. Yet, it is important to consider that longer-term trends and historical context can also contribute to understanding the dynamics of energy poverty. The results of our study may be combined with earlier studies or historical data to gain a more comprehensive understanding of the long-term impact of energy poverty on respiratory infections in children in LMICs.

    Conclusions

    Multidimensional energy poverty is associated with greater odds of ARI in children under 5 years of age living in 22 LMICs. In particular, using biomass for cooking and lack of access to electricity and entertainment, education and household appliances increased the odds of ARI. These findings justify the design and implementation of interventions that address energy poverty from a multidimensional perspective and improve both energy affordability and accessibility. Causal relationships between multidimensional energy poverty and child ARI in LMICs require further investigation.

    Data availability statement

    Data are available in a public, open access repository. All data relevant to the study are included in the article or uploaded as online supplemental information. All data relevant to the study are included in the article or available as online supplemental material. The data used in the analyses are publicly available, anonymised and geographically scrambled to ensure confidentiality. More information on DHS can be found at 'https://dhsprogram.com/ (accessed on 8 August 2022)', where survey datasets can be obtained.

    Ethics statements

    Patient consent for publication

    Ethics approval

    Procedures and surveys relevant to the data used in this study have been reviewed and approved by International Coaching Federation (ICF) Institutional Review Board. ICF ensures that the surveys comply with the Human Subjects Protection Act of the US Department of Health and Human Services. Permission and data access was obtained from The DHS Programme for this study. Our secondary analyses of the data were deemed exempt from additional review according to the Dutch Medical Research with Human Subjects Law by the Erasmus Medical Center ethical review committee (ref. no. MEC-2022-0560).

    References

    1. ↵
      2. Pan L ,
      3. Biru A ,
      4. Lettu S
      . Energy poverty and public health: global evidence. Energy Economics 2021;101:105423. doi:10.1016/j.eneco.2021.105423
      OpenUrl
    2. ↵
      2. Ballesteros-Arjona V ,
      3. Oliveras L ,
      4. Bolívar Muñoz J , et al
      . What are the effects of energy poverty and interventions to ameliorate it on people’s health and well-being?: a scoping review with an equity lens. Energy Research & Social Science 2022;87:102456. doi:10.1016/j.erss.2021.102456
      OpenUrl
    3. ↵
      1. United Nations
      . Transforming our world: The 2030 agenda for sustainable development. New York: UN Publishing, 2015.
    4. ↵
      2. Zhang D ,
      3. Li J ,
      4. Han P
      . A multidimensional measure of energy poverty in China and its impacts on health: an empirical study based on the China family panel studies. Energy Policy 2019;131:72–81. doi:10.1016/j.enpol.2019.04.037
      OpenUrl
    5. ↵
      2. Nussbaumer P ,
      3. Bazilian M ,
      4. Modi V
      . Measuring energy poverty: focusing on what matters. Renewable and Sustainable Energy Reviews 2012;16:231–43. doi:10.1016/j.rser.2011.07.150
      OpenUrl
    6. ↵
      1. World Health Organization (WHO)
      . Household air pollution and health. 2022. Available: https://www.who.int/news-room/fact-sheets/detail/household-air-pollution-and-health [Accessed 26 Sep 2022].
    7. ↵
      1. IEA, IRENA, UNSD, World Bank, WHO
      . Tracking SDG 7: the energy progress report. Washington DC: World Bank, 2022.
    8. ↵
      2. González-Eguino M
      . Energy poverty: an overview. Renewable and Sustainable Energy Reviews 2015;47:377–85. doi:10.1016/j.rser.2015.03.013
      OpenUrl
    9. ↵
      2. Simoes EAF ,
      3. Cherian T ,
      4. Chow J , et al
      . Acute respiratory infections in children. 2006.
    10. ↵
      2. Woolley KE ,
      3. Bartington SE ,
      4. Kabera T , et al
      . Comparison of respiratory health impacts associated with wood and charcoal biomass fuels: a population-based analysis of 475,000 children from 30 Low- and middle-income countries. Int J Environ Res Public Health 2021;18:9305. doi:10.3390/ijerph18179305
    11. ↵
      2. Lee KK ,
      3. Bing R ,
      4. Kiang J , et al
      . Adverse health effects associated with household air pollution: a systematic review, meta-analysis, and burden estimation study. Lancet Glob Health 2020;8:e1427–34. doi:10.1016/S2214-109X(20)30343-0
      OpenUrl
    12. ↵
      2. Gordon SB ,
      3. Bruce NG ,
      4. Grigg J , et al
      . Respiratory risks from household air pollution in low and middle income countries. Lancet Respir Med 2014;2:823–60. doi:10.1016/S2213-2600(14)70168-7
      OpenUrl
    13. ↵
      2. Yaya S ,
      3. Bishwajit G
      . Burden of acute respiratory infections among under-five children in relation to household wealth and socioeconomic status in Bangladesh. Trop Med Infect Dis 2019;4:36. doi:10.3390/tropicalmed4010036
    14. ↵
      2. Merera AM
      . Determinants of acute respiratory infection among under-five children in rural Ethiopia. BMC Infect Dis 2021;21:1203. doi:10.1186/s12879-021-06864-4
    15. ↵
      1. The Demographic and Health Survey Program (DHS)
      . The DHS program US-AID. 2012. Available: https://dhsprogram.com/ [Accessed 8 Aug 2022].
    16. ↵
      2. von Elm E ,
      3. Altman DG ,
      4. Egger M , et al
      . The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol 2008;61:344–9. doi:10.1016/j.jclinepi.2007.11.008
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      2. Boyle EH ,
      3. King M ,
      4. Sobek M
      . Data from: IPUMS-demographic and health surveys: version 9. IPUMS and ICF, 2022. doi:10.18128/D080.V9
    18. ↵
      1. The World Bank
      . World Bank country and lending groups [online]. Available: https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bankcountry-and-lending-groups [Accessed 9 Aug 2022].
    19. ↵
      1. UNICEF
      . Multiple indicator cluster surveys. 2019. Available: https://mics.unicef.org/ [Accessed 26 Sep 2022].
    20. ↵
      2. Rehfuess EA ,
      3. Best N ,
      4. Briggs DJ , et al
      . Diagram-based analysis of causal systems (DACS): elucidating inter-relationships between determinants of acute lower respiratory infections among children in sub-Saharan Africa. Emerg Themes Epidemiol 2013;10:13. doi:10.1186/1742-7622-10-13
    21. ↵
      2. Troeger CE ,
      3. Khalil IA ,
      4. Blacker BF , et al
      . Quantifying risks and interventions that have affected the burden of lower respiratory infections among children younger than 5 years: an analysis for the Global Burden of Disease Study 2017. The Lancet Infectious Diseases 2020;20:60–79. doi:10.1016/S1473-3099(19)30410-4
      OpenUrlPubMed
    22. ↵
      2. Rehfuess EA ,
      3. Briggs DJ ,
      4. Joffe M , et al
      . Bayesian modelling of household solid fuel use: insights towards designing effective interventions to promote fuel switching in Africa. Environ Res 2010;110:725–32. doi:10.1016/j.envres.2010.07.006
      OpenUrlPubMed
    23. ↵
      1. United Nations Human Settlements Programme
      . Urban indicators guidelines: monitoring the habitat agenda and the Millennium Development Goals. Nairobi: UN Habitat, 2004.
    24. ↵
      1. World Bank Group
      . Climate change knowledge portal [online]. Available: https://climateknowledgeportal.worldbank.org/ [Accessed 29 Nov 2022].
    25. ↵
      2. Adesanya OA ,
      3. Chiao C
      . Environmental risks associated with symptoms of acute respiratory infection among preschool children in North-Western and South-Southern Nigeria communities. Int J Environ Res Public Health 2017;14:1396. doi:10.3390/ijerph14111396
    26. ↵
      2. Buchner H ,
      3. Rehfuess EA
      . Cooking and season as risk factors for acute lower respiratory infections in African children: a cross-sectional multi-country analysis. PLOS One 2015;10:e0128933. doi:10.1371/journal.pone.0128933
    27. ↵
      2. Odo DB ,
      3. Yang IA ,
      4. Dey S , et al
      . Ambient air pollution and acute respiratory infection in children aged under 5 years living in 35 developing countries. Environ Int 2022;159:107019. doi:10.1016/j.envint.2021.107019
      OpenUrl
    28. ↵
      2. Terfa ZG ,
      3. Ahmed S ,
      4. Khan J , et al
      . Household microenvironment and under-fives health outcomes in Uganda: focusing on multidimensional energy poverty and women empowerment indices. Int J Environ Res Public Health 2022;19:6684. doi:10.3390/ijerph19116684
    29. ↵
      1. R Core Team
      . R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2022. Available: https://www.R-project.org/
    30. ↵
      1. IBM Corp
      . IBM SPSS statistics for windows, version 28.0. Armonk, New York: IBM Corp, 2021.
    31. ↵
      2. Hasan MM ,
      3. Richardson A
      . How sustainable household environment and knowledge of healthy practices relate to childhood morbidity in South Asia: analysis of survey data from Bangladesh, Nepal and Pakistan. BMJ Open 2017;7:e015019. doi:10.1136/bmjopen-2016-015019
    32. ↵
      2. Irwin BR ,
      3. Hoxha K ,
      4. Grépin KA
      . Conceptualising the effect of access to electricity on health in low- and middle-income countries: a systematic review. Glob Public Health 2020;15:452–73. doi:10.1080/17441692.2019.1695873
      OpenUrl
    33. ↵
      1. World Health Organization
      . Air pollution and child health: prescribing clean air. Summary (WHO/CED/PHE/18.01). Licence: CC BY-NC-SA 3.0 IGO. Geneva: World Health Organization, 2018.
    34. ↵
      2. Dherani M ,
      3. Pope D ,
      4. Mascarenhas M , et al
      . Indoor air pollution from unprocessed solid fuel use and pneumonia risk in children aged under five years: a systematic review and meta-analysis. Bull World Health Organ 2008;86:390–398C. doi:10.2471/blt.07.044529
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      2. Winskill P ,
      3. Hogan AB ,
      4. Thwing J , et al
      . Health inequities and clustering of fever, acute respiratory infection, diarrhoea and wasting in children under five in Low- and middle-income countries: a demographic and health surveys analysis. BMC Med 2021;19:144. doi:10.1186/s12916-021-02018-0
    36. ↵
      1. IEA
      . World energy outlook 2019. Paris: IEA, 2019.
    37. ↵
      2. Quansah R ,
      3. Semple S ,
      4. Ochieng CA , et al
      . Effectiveness of interventions to reduce household air pollution and/or improve health in homes using solid fuel in low-and-middle income countries: a systematic review and meta-analysis. Environ Int 2017;103:73–90. doi:10.1016/j.envint.2017.03.010
      OpenUrlCrossRefPubMed
    38. ↵
      2. Lantz B
      . The large sample size fallacy. Scand J Caring Sci 2013;27:487–92. doi:10.1111/j.1471-6712.2012.01052.x
      OpenUrl
    39. ↵
      2. Diez-Roux AV
      . Multilevel analysis in public health research. Annu Rev Public Health 2000;21:171–92. doi:10.1146/annurev.publhealth.21.1.171
      OpenUrlCrossRefPubMedWeb of Science
      2. Abbas K ,
      3. Xie X ,
      4. Xu D , et al
      . Assessing an empirical relationship between energy poverty and domestic health issues: a multidimensional approach. Energy 2021;221:119774. doi:10.1016/j.energy.2021.119774
      OpenUrl
      2. Crentsil AO ,
      3. Asuman D ,
      4. Fenny AP
      . Assessing the determinants and drivers of multidimensional energy poverty in Ghana. Energy Policy 2019;133:110884. doi:10.1016/j.enpol.2019.110884
      OpenUrl

    Supplementary materials

    Footnotes

    • Contributors MS initiated the study and proposed the research question. MS and AvG designed the study. MS performed the statistical analysis and designed tables and figures. DN, AvG, JY-H and SZ aided in the statistical analysis. MS wrote the first draft of the manuscript. All authors critically and equally reviewed and edited the manuscript. MS is the guarantor.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.