Correlates of Ghanaian households’ use of clean cooking fuels: insight from the 2014 demographic and health surveys

Background: Globally, about 2.9 billion persons, the majority of whom are from low and middle-income countries, depend on poorly combusted biomass and coal for cooking and heating. Given the negative toll the use of poorly combusted fuels have on people’s health and quality of life, homes are admonished to use for cooking clean fuels that promote optimal health such as liqueed petroleum gas (LPG), natural gases, electricity, and improved biomass cook-stoves. The use of clean cooking fuels has other co-benets such as reduce deforestation and lower climate-changing emissions. Although information on the prevalence of access to clean cooking energy is known, studies on the social, economic, and geographic factors that may explain the low adoption in households are rare. In this paper, we aim to determine the major covariates of households’ access to clean cooking energy in Ghana. Methods: The study used the data of the cross-sectional demographic and health surveys conducted in 2014. Chi-square test of independence was used to identify the covariates that were statistically signicant related to households’ use of CF&HPFs for the multivariable model. In the multivariable model, we reported prevalence ratios using the generalized linear model (glm), setting the family to “Poisson” and the link to “log.” The estimation adjusted for the study designed and the household sampling weights. Results: The main nding was that rich households were 8.85 times likely to use CF&HPFs compared to poor households. We further found that the following factors were associated with a higher probability of a Ghanaian household to use climate-friendly and health-promoting fuels for cooking: male-headed households, households’ headed by persons between 24years and 55 years, households with heads that were currently married or never married, households with heads that have attained at least a primary level education, households with less than eight members, households with at least two women, and households in Western, Central, Greater Accra, Volta, Eastern, Ashanti, Brong Ahafo, Upper East region. Conclusion: In view of these ndings, we recommend: a national effort to subsidize and make affordable the purchase of LPG and the stoves that burn it even to the poorest household; education on

These fuel types emit health and climate destructive pollutants such as carbon dioxide, carbon monoxide, benzene, butadiene, nitrogen oxides, formaldehyde, polynuclear aromatic hydrocarbons, and ne particulate matter [3,4]. Globally, household cooking and heating account for 58% of global black carbon emissions [5].
Prolonged exposure to the by-products of poorly combusted fuels is associated with unimaginably high morbidity and mortality records among the aged, adults, and children [1,2,6]. About 4 million persons, 54% of whom are women and children, lose their lives annually from using poorly combusted fuels [2].
Incomplete combusted products can adversely affect household air quality, resulting in a myriad of adverse health effects among children and women [1,2,6]. This polluted home environment is associated with acute lower respiratory infections such as pneumonia and bronchitis in children [1,6].
This environment equally contributes to increased risk of chronic obstructive pulmonary disease and various types of cancer, and poor pregnancy outcomes such as low birth weight and neonatal deaths among women [1].
The use of traditional biomass fuels poses several threats to the ecosystem and the protection they provide [7]. The human activity of tree harvesting for fuel leads to forest degradation with extended negative effects on biodiversity, soil stability, and water quality and quantity [7]. The production and use of charcoal for household cooking, for instance, produces climate-altering pollutants (CAPs) such as black carbon and carbon dioxide [8]. The protection trees and green vegetation provide is that they absorb atmospheric carbon dioxide, the prominent CAP, for photosynthesis [7]. Therefore, the harvesting of trees weakens nature's ability to effectively participate in the carbon cycle to the bene t of both the environment and human health [7,8].
Given the negative toll the use of poorly combusted fuels have on people's health and quality of life, homes are admonished to use for cooking and heating clean fuels that promote optimal health such as lique ed petroleum gas (LPG), natural gases, electricity, and in some cases improved biomass cookstoves [2,4,9]. The use of clean cooking energy sources has other co-bene ts such as reduce deforestation and lower climate-changing emissions [2,4]. Although LPG and natural gas are fossil and burning them emit carbon, their net contribution to the global black carbon and impact on the environment are lesser compared to the burning of biomass fuels [9]. Given that electricity's climate friendliness depends on the energy sources, many countries including Ghana are making efforts to generate their electricity by using fuels that pose minimal threat to the environment. Ghana, for instance, has a total of 2520mw installed electricity generation capacity. About 47% and 13% of the electricity is entirely generated by hydropower and gas, respectively [10]. The remaining 40% is interchangeably generated through both gas and oil [10]. Therefore, the push by the International Energy Agency, the United Nations (UN), and the World Health Organization (WHO) to universally adopt these types of fuels as clean cooking solutions is in the interest of promoting human health and saving the climate [2] .
Despite the progress made by governments, non-governmental organizations, and energy advocates to use clean fuels to meet household energy needs, the current estimate suggests that 4 out of 10 people worldwide do not have access to clean cooking energy sources [2]. Sub-Sahara Africa (SSA) has the largest clean cooking energy access de cit with only about 14% of its population having access. Although Ghana's estimate of 25% of the population having access to clean cooking energy sources is higher than the SSA average [2], it is undesirable and puts Ghana in an impossible position of achieving the SDG 7 of universal adoption of clean cooking energy by 2030.
Although information on the prevalence of access to clean cooking energy is known, studies on the social, economic, and geographic factors that may explain the low adoption in households are rare. In this paper, we aim to determine the major covariates of households' access to clean cooking energy in Ghana.

Design
The 2014 Ghana Demographic and Health Surveys (GDHS) data was collected in line with a crosssectional survey design [11].The survey employed two-stage probability sampling, strati ed by urban and rural areas of the ten administrative regions of Ghana [11]. The sampling frame for the survey was based on the updated sampling frame from the 2010 Ghana Population and Housing Census [11]. The rst stage of the probability sampling involved the selection of enumeration areas (EAs), which was strati ed by place of residence [11]. A total of 427 EAs were selected, 216 in urban areas and 211 in rural areas [11]. The second stage of the probability sampling involved the systematic sampling of households [11].
The implementers of the survey undertook a household-listing operation in all the selected EAs from January through March 2014, and households to be included in the survey were systematically selected from the list [11]. Approximately 30 households were selected from each EA to constitute the total sample size of 12,831 households [11].

Data collection
Trained enumerators collected the data from early September to mid-December 2014 using paper-based questionnaires [11]. The selected sample size for the 2014 GDHS was 12,831 households, of which 12,010 were occupied [11]. Out of the occupied households, 11,835 were successfully reached, resulting in a 99% response rate [11]. In the GDHS, household heads provided information on their demographic characteristics and household characteristics such as household population and composition, housing structure, household assets, access to basic utilities, sources of drinking water, water treatment practices, access to sanitation facilities, and type of fuel used for cooking [11].

Study sample
The unit of analysis is households, and the dataset contains 11,835 households. The data were weighted using the household weight variable in the dataset. A sample of 513 households indicated that they cooked no food in the house, so they were excluded from further analyses. Therefore, the analytic weighted sample is 11,322.

Variables
Outcome variable: In the dataset, the household head was asked to select the main source of fuel for cooking in the household from the following categories: electricity, LPG, natural gas, kerosene, coal, lignite, charcoal, wood, straw/shrubs/grass, and agricultural crop. Electricity, LPG, and natural gas was de ned in our study as HP & CFFs and all other sources of fuels as non-climate friendly and health promoting.
Additional le shows the proportion of households using each of the fuel types for cooking in Ghana (See Additional le 1). Covariates: Ten sociodemographic characteristics variables were selected as potential covariates in the study. Marital status was recoded as follows: never in union (as never married), married and living with partner (as currently married), widowed, divorced, and no longer living together/separated (as formally married). For the multivariable model, household wealth was dichotomized as follows: poorest, poorer, and middle (as poor) and richer and richest (as rich).
Household wealth index was already estimated and reported in the DHS data. This was created using household characteristics (source of drinking water, type of toilet, sharing of toilet facilities, main material for roof, walls and oors oor, and type of cooking fuel amongst others household characteristics) and household possessions and assets (ownership of television, radio, vehicle, bicycles, motorcycles, watch, agricultural land, farm animals/livestock, and bank account amongst others). DHS used a principal component analysis (PCA) to assign weights to each asset in each household and cumulative score were calculated from the assigned weights. Households were ranked according to the cumulative scores from the household assets. The cumulative percentage distribution of the wealth score was estimated and the wealth score values that corresponded to the four cut point values of the quintiles (20th, 40th, 60th, and 80th percentiles) were determined. Households with values less than or equal to the 20th percentile score were assigned poorest, those greater 20% but less than or equal to 40th % were assigned poorer, those greater than the 40th % and less than or equal to the 60th % score were assigned middle, greater than the 60th % and less than or equal to the 80th % score were assigned richer and the richest household were those with scores greater than the 80th percentile score. Wealth index was thus ranked into quintiles: poorest, poorer, middle, richer, and richest [12].

Analytic procedure
We used STATA version 14 for the data analyses. Summary statistics, chi-square test of independence, and Poisson regression were performed. The analysis accounted for sampling design (cluster, strata) and household weights using the 'svy' command in a default mode. The default 'svy' computes standard errors by using the linearized variance estimator called the rst-order Taylor linearization. This process eliminated the incorrect estimation of the Standard Errors (SE) associated with the con dence intervals of the regression coe cients. After survey setting our data, we obtained prevalence ratios by performing Poisson regression. We achieved this by using the generalized linear model (glm) in STATA, setting the family to "Poisson" and the link to "log," to avoid the potential overestimation of effects from reporting odds ratios when the prevalence of the outcome of interest is above 10%. The mathematical equation for Poisson regression is as follows: (see Formulas in the Supplementary Files) The outcome variable has only two events (that is '0, 1'), where the value '1' means a household was using clean cooking fuel and the value '0' means that a household is not using clean cooking fuel.
Before performing the multivariable analysis, we check the assumption of multicollinearity and no violations were observed. The variance in ation factors of the covariates are all less than 2 (Table 1), which are far less than the cut of points of 10 [13].  Chi-square test of independence between the outcome variable and potential covariates Results from the chi-square test of independence revealed that sex, age, marital status, and education of HH, size of household, number of women or men in the household, household wealth, locality of residence, and region of residence were statistically signi cantly associated with the use of HP & CFFs for cooking. Covariates of climate-friendly and health-promoting cooking fuels We built a robust Poisson model of covariates of health-promoting and climate-friendly cooking fuels with variables that were statistically signi cantly related to the outcome variable in the bivariate analysis.
The following factors were associated with a higher probability of a Ghanaian household to use HP & CFFs for cooking: male-headed households, households' headed by persons between 24 years and 55 years, households with heads that were currently married or never married, households with heads that have attained at least a primary level education, households with less than eight members, households with at least two women, rich households, and households in Western, Central, Greater Accra, Volta, Eastern, Ashanti, Brong Ahafo, Upper East region. Details of the prevalence ratio estimates of each covariate in the model with its corresponding 95% con dence intervals are reported in Table 4.

Discussion
The study sets out to estimate the proportion of Ghanaian households currently using CF&HPFs and delineate factors that account for their use.
The ndings reveal that less than a third of Ghanaian households were using CF&HPFs for cooking. A decade ago, the proportion of households that had access to LPG was estimated to be 18% [14]. Our result show that Ghana's progress regarding access to clean cooking fuels has been slow. Since the early 1990, the government of Ghana has implemented Lique ed Petroleum Gas (LPG) program to encourage households to use (LPG) for cooking as a better alternative to the commonly used traditional biomass fuels [15]. The program has mainly undertaken safety awareness campaign, local production of cylinder and its accessories and enabling LPG accessibility to the population by distributing free LPG gas cylinders [15]. The LPG cylinder distribution was further intensi ed in 2013. All these programs had an objective of making 50% of Ghanaian households to use LPG for household energy need by 2015 [16], 2020 (revised timeline) [14], and 2030 (new timeline) [17]. The revision of the timelines were due to Ghana's LPG promotion program failure to meet its target. Given the current estimate and the slow adoption rate for clean cooking solutions, it will be prudent for the government of Ghana to review and evaluate her LPG promotion programs since the 1990.
The households' use of CF&HPFs were also found to be positively associated with certain household head demographic factors, household characteristics, and place and region of residence. To begin with, male-headed households were more likely to use CF&HPFs for cooking compared to female-headed households. Men have more economic advantage in most Ghanaian communities. This advantage position of men enables them to purchase CF&HPFs that are generally considered expensive in Ghana. Even though some studies present similar results, others present contrary ndings. For example, one study conducted in South Asia found that female-headed households were more likely to use CF&HPFs fuels [18,19]. The authors mentioned the ability of CF&HPFs to reduce the drudgery associated with cooking as a major in uential factor for its adoption and use by female-head households.
Again, households headed by persons between 24 years and 55 years were observed to more likely to use CF&HPFs. Conventionally, persons within this age group belong to the active working-age population, granting them easier access to economic opportunities. Therefore, these individuals are more likely to afford the usage of CF&HPFs. Additionally, persons above 55 years may nd it di cult and inconvenient to adopt CF&HPFs for cooking because they perhaps belonged to a generation that predominantly relied on solid fuels for cooking [20]. Other scholars also explain that the tendency to experiment and to adopt e cient modern technologies and behaviour may be a contributing factor for the adoptions and use of CF&HPFs among the younger generation [21]. Furthermore, we found that households with currently married or never-married individuals were more likely to use CF&HPFs. In today's Ghana, it is trendy for couples to rely on LPG and electricity for cooking.
Also, anecdotal evidence suggests that couples may be more likely to use CF&HPFs because of its convenience. Even though undocumented in the literature, it is prudent to argue that households with never-married heads (majority of whom are young adults) were more likely to use CF&HPFs due to its convenience and the high tendency of reducing exertion often associated with traditional cooking methods.
Moreover, we found education as another key contributing factor in uencing adoption and use of CF&HPFs among households. This is because the educated household head is more informed on the environmental and health bene ts of cleaner cooking fuels, which is likely to in uence its choice and usage. One scholar in a similar study also opined that educated household heads in urban communities are more inclined to use CF&HPFs [22]. In addition, it is instructive to reason that educated household heads may have more access to economically viable opportunities than their uneducated counterparts.
This economic advantage provides educated households with high-income levels that enable them to purchase CF&HPFs which are relatively expensive but convenient to use.
Besides, the study revealed that household's with less than eight members have a higher likelihood of using CF&HPFs. We attribute this phenomenon to the reduction in the nancial burden in smaller households than larger households. Large households in Ghana are most often dominated by dependent individuals. The high demand for cooking energy uses due to the family size and the challenge of limited resources create the situation of treating the adoption and use of CF&HPFs which are most often expensive as a periphery or less of a household need due to prevailing household welfare status. Other studies con rms this nding [23][24][25].
We also found that households with at least two women had a higher likelihood of using CF&HPFs. In Ghana, cooking activities are mostly a preserve of adult female household members. This exposes them, (especially women without clean cooking methods) to chronic obstructive pulmonary diseases (COPD) such as bronchitis, lung cancer, asthma etc than their counterparts who adopt and use CF&HPFs. We argue that in an attempt to avert the physical drain and negative health implications associated with the burning of solid biomass for cooking, households with 2 or more adult females are more likely to combine nancial resources to adopt CF&HPFs which is more convenient, timesaving and healthy.
Findings from a study conducted by some authors concur with this observation [25].
Our study revealed that rich households were 8.85 times likely to use CF&HPFs compared to poor households. The adoption and use of cooking fuels like other graded commodities are highly in uenced by the purchasing power of the household. The nding agrees with the Energy Ladder Theory that explains the direct relationship between the rise in household income and the adoption of clean cooking energy sources [26]. The mean annual non-food and housing expenditure for households in Ghana is GHC 4,682 [27]. Averagely, 14.5kg of LPG should su ce to meet a household's cooking and heating need.
The price for 14.5kg of LPG is about GHC 80/USD 13.8 (GHC 960/USD 166 annually), representing about 21% of annual non-food-and-housing expenditure. The average annual gross income for poor households is GHC 18,191, which is 2.5 times lower than that of the rich household (GHC 44, 933) [27]. Arguably, allocating one-fth of non-food-and-housing expenditure to clean cooking fuel will be much easier for the rich households than the poor. Again, apart from the ability to purchase CF&HPFs, there appears to be some luxurious prestige attached with the adoption and use of CF&HPFs which rich households want to be associated with [24].
We investigated whether regional variations in uenced the likelihood of a Ghanaian household to use climate-friendly and health-promoting fuels for cooking. The study observed that households within some political administrations in Ghana had a higher probability of using improved cooking methods than others. For example, we observed that households in the Greater Accra, Ashanti, Brong Ahafo, Central, Western and Eastern regions were more likely to use CF&HPFs than traditional biomass. Possible explanation for this observation can be attributed to the urbanized nature of these regions and availability and distribution of LPG gas re lling stations [28]. The Greater Accra region, for example, is the political head and administrative region of Ghana. It is the most urbanized region with about 87.4% of its population living in urban centres with access to viable economic and political resources for an improved standard of living [29]. This implies that households within these regions are more likely to utilize the available economic and political advantage to adopt or adapt climate-friendly and health-promoting fuels for cooking than their counterparts in the other regions of the country. Secondly, the reference region (Northern) has only twelve LPG gas re lling stations compared to Greater Accra (179 outlets), Ashanti (95 outlets), Brong Ahafo (70 outlets), Central (76 outlets), Western (66 outlets) and Eastern (66 outlets) regions which have over 65 re lling outlets [30], making it easier for households in these regions to have access to LPG for their household needs. These nding shares some similarities with ndings of one other study [31].
Of particular interest in the above nding was the higher probability of households in the Upper East region to use climate and health-promoting fuels for cooking. This is quite revealing because households in the Northern regions of Ghana are the poorest across the country and with lower number of LPG re lling outlets. However, the likelihood of households in the Upper East regions identi ed in this study to use CF&HPFs suggests some qualitative explanations. In 2013, Ghana launched The Rural LPG Promotion Program (RLP) in the Upper East Region with the aim to create access to households in rural areas to use LPG for cooking [30]. The RLP involved the distribution of free gas cylinders to households [30]. It is therefore prudent to argue that many households in the region had a higher chance to bene t from the free cylinders compared with the Northern region.
The paper has certain strength and limitations that are worth mentioning. The study employed the crosssectional analysis which did not permit the researchers to derive causal relationships between covariates and the outcome variable. It is also worth mentioning that few households were excluded from the analysis because they did not cook at home and therefore did not use any domestic cooking energy. Close to four decades, the GDHS program has been conducting its survey every ve years, and this has improved the quality of the sampling methodologies, data collection procedures, and the development of the questionnaires. Given that the study used the GDHS dataset that is nationally representative and large, the estimates from the analyses can be generalized for the entire Ghanaian population.

Conclusions
The adoption and sustained use of CF&HPFs are important in enhancing good health for nearly half of the world's population, mostly women and children in developing countries and to help protect the global environment. Our study nds household characteristics and region of residence as in uential factors in the adoption of CF&HPFs in Ghana. The level of education, wealth status, age and marital status of household heads positively in uenced the use of CF&HPFs. It became evident in this study that maleheaded households were more likely than female-headed households to adopt CF&HPFs even though it differed from most literature. Additionally, having two or more adult females in a household positively in uences the chances of using clean cooking energy. Large family size (above 8 individuals) and residence in the Upper West region of Ghana were found to negatively in uence the adoption of clean cooking fuels.
To increase the adoption and use of CF&HPFs to advance Ghana's cause towards achieving the SDGs, we recommend: a national effort to subsidize and make affordable the purchase of LPG and the stoves that burn it even to the poorest household; education on e cient and sustainable consumption must be intensi ed and incorporated in the basic school's curriculum to ensure the future generation are more environmental sustainability and health-conscious; creation of multiple LPG re lling outlets in all the regions in Ghana.

Declarations
Ethics approval and consent to participate The Ghana Health Service Ethical Review Committee and the Institutional Review Board of ICF International reviewed and approved the 2014 GDHS protocol. Enumerators obtained informed consent from mothers between 15-49 years with children under-ve years before interviewing them [11]. The GDHS did not obtain assent from parents or guardians of childbearing women who were either 15 or 16 years because they already had children and were independent of their guardians or parents. The GDHS is publicly available upon a simple, registration-access request, so no further ethical clearance was sought Consent for publication

Not Applicable
Availability of data and materials The 2014 Ghana DHS dataset supporting the analysis of this study is available in the DHS repository. The DHS datasets are available for free after a simple registration process.

Competing Interests
The authors declare no competing interests.

Funding
Authors did not receive any funding.
Authors' contributions DSS, AP and PLK conceptualized the study. AP and DOH performed data analysis. NKE validated the analysis. DSS, PLK, NKE, DMV, ANKH, DOH and AP wrote the paper. AP supervised the study. All authors read and approved the nal manuscript.