Comprehensive health risk assessment of urban ambient air pollution
(PM2.5, NO2 and O3) in Ghana

Boansi Adu Ababio a,e,*, Gerheart Winfred Ashong a, Thomas Peprah Agyekum b,
Blessed Adjei Yeboah c , Marian Asantewah Nkansah a, Jonathan Nartey Hogarh d,
Michael Kweku Commeh e, Edward Ebow Kwaansa-Ansah a, Kwabena Dabie f, Felix Adulley f,
Eldad Boansi e, Lorenda Sarbeng g, Birago Adu Ababio h, Maame Serwaa Boapea i,
Nana Kwabena Oduro Darko j, Meshach Kojo Appiah k

a Department of Chemistry, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana
b Department of Occupational & Environmental Health & Safety, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana
c Department of Computer Science, Kumasi Technical University, Kumasi, Ghana
d Department of Environmental Science, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana
e Technology Consultancy Centre International Centre for Innovation, Manufacturing, Technology Transfer and Entrepreneurship, Kwame Nkrumah University of Science
and Technology, Kumasi, Ghana
f Department of Biochemistry, University of Cape Coast, Cape Coast, Ghana
g Department of Geography and Regional Planning, University of Cape Coast, Cape Coast, Ghana
h Department of Biomedical Sciences, University of Health and Allied Sciences, Ho, Ghana
i Department of Virology, Noguchi Memorial Institute for Medical Research, University of Ghana, Legon, Ghana
j Department of Biochemistry and Biotechnology, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana
k Department of Basic Education, University of Education, Winneba, Ghana

A R T I C L E I N F O

Edited by Professor Bing Yan

Keywords:
Air pollution
Health risk
Air quality
Ghana
Particulate matter
Nitrogen dioxide
Ozone

A B S T R A C T

Urbanization and industrialization have drastically increased ambient air pollution in urban areas globally from
vehicle emissions, solid fuel combustion and industrial activities leading to some of the worst air quality con-
ditions. Air pollution in Ghana causes approximately 28,000 premature deaths and disabilities annually, ranking
as a leading cause of mortality and disability-adjusted life years. This study evaluated the annual concentrations
of PM2.5, NO2 and O3 in the ambient air of 57 cities in Ghana for two decades using historical and forecasted data
from satellite measurements. The study assessed urban air quality and evaluated both carcinogenic and non-
carcinogenic health risks associated with human exposure to ambient air pollutants. Alarmingly, our findings
revealed the yearly median PM2.5 concentrations (50.79–67.97 µg m− 3) to be significantly higher than the WHO
recommendation of 5 µg m− 3. Tropospheric ozone concentrations (72.21–92.58 µg m− 3 ) also exceeded the WHO
annual standard of 60 µg m− 3. Furthermore, NO2 concentrations (3.65–12.15 µg m− 3 ) surpassed the WHO
threshold of 10 µg/m³ in multiple cities. Hazard indices indicated that PM2.5 and O3 pose significant non-
carcinogenic health risks for younger age groups for a daily exposure duration of three hours and beyond. Ac-
cording to the Air Quality Life Index (AQLI) in our study, exposure to PM2.5 shortens life expectancy by 4.5–6.2
years. The ambient air of the majority (98 %) of the cities was unhealthy for sensitive groups. This study reveals
the urgent need for comprehensive air quality policies in Ghanaian cities. It emphasizes the significance of robust
real-time monitoring of air pollutants and the investigation of seasonal dust storm effects, to fill data gaps in
Ghana and West Africa, facilitating evidence-based interventions that improve urban air quality and public
health outcomes.

* Corresponding author at: Department of Chemistry, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana.
E-mail address: boansiaduababio@gmail.com (B.A. Ababio).

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

https://doi.org/10.1016/j.ecoenv.2024.117591
Received 2 August 2024; Received in revised form 16 December 2024; Accepted 19 December 2024

Ecotoxicology and Environmental Safety 289 (2025) 117591

Available online 7 January 2025
0147-6513/© 2024 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-
nc-nd/4.0/ ).

https://orcid.org/0009-0005-3133-9271
https://orcid.org/0009-0005-3133-9271
mailto:boansiaduababio@gmail.com
www.sciencedirect.com/science/journal/01476513
https://www.elsevier.com/locate/ecoenv
https://doi.org/10.1016/j.ecoenv.2024.117591
https://doi.org/10.1016/j.ecoenv.2024.117591
http://crossmark.crossref.org/dialog/?doi=10.1016/j.ecoenv.2024.117591&domain=pdf
http://creativecommons.org/licenses/by-nc-nd/4.0/
http://creativecommons.org/licenses/by-nc-nd/4.0/

1. Introduction

Air is a mixture of gases and tiny dust particles. It is a basic element
of life, as a source of oxygen needed for respiration. It is a microbial
habitat and a superhighway for the water cycle (Ababio, 2023). Air
pollution is caused by human activities or natural processes that intro-
duce harmful substances into the air, posing risks to the health and
well-being of humans and endangering the ecosystem at large. Ambient
air pollution is a major health risk in the 21st century. It is responsible
for millions of annual premature mortalities across the globe in urban
and rural environments (Li et al., 2023). Half of the world’s population
resides in urban cities where they are exposed to gaseous and particulate
pollutants (Fan et al., 2020). It is projected that by 2050 approximately
70 % of the global population will live in urban areas (Sicard et al.,
2023). Nitrogen dioxide (NO2), particulates with aerodynamic sizes of
2.5 µm (PM2.5) and tropospheric ozone (O3) are among the most harmful
air pollutants to public health in urban settings (Sicard et al., 2020;
Cakaj et al., 2023). Millions of people are exposed to the concentrations
of these pollutants at levels beyond the WHO-recommended thresholds
(Anenberg et al., 2022; Sicard et al., 2021; Southerland et al., 2022).
Chronic exposure to these pollutants results in increased risks of
ischaemic heart disease, stroke, acute lower respiratory infections,
chronic obstructive pulmonary diseases, acute kidney injury, dementia,
cognitive decline, liver and lung cancer (Kumar et al., 2023; Wang et al.,
2023; Ma et al., 2023; 2024).

Urbanization and industrialization have led to a significant increase
in ambient air pollution levels in urban cities across the globe making
them have some of the worst air quality on the planet. Anthropogenic
sources of air pollution comprise vehicle emissions, the combustion of
solid fuels and domestic waste as well as industrial, agricultural, con-
struction and mining activities (Nyarku et al., 2021; Okedere et al.,
2021). The Global Burden of Disease estimates ambient air pollution to
be a primary contributor to global mortality and the loss of
disability-adjusted life years, with 4 million premature deaths annually
linked to it (Johnson et al., 2024). Air pollution is the second-highest
risk factor in Ghana, accounting for about 28,000 premature deaths
and disabilities per annum (Ababio et al., 2023). The World Bank esti-
mates that the annual cost of air pollution in Ghana is 4.2 % of GDP,
which is equivalent to US $2.5 billion. Also, it costs approximately US
$264 million per annum alone for the largest urban cities, in Ghana,
such as Accra and Kumasi (World Bank Group, 2020).

Various kinds of studies including monitoring and measurement
studies, epidemiological investigations, source apportionment analyses,
air quality modeling, policy and regulatory assessments, control and
mitigation studies have been conducted on air pollution in the ambient
environment across the globe. Despite these numerous studies, the state
of ambient air pollution in urban cities in Sub-Saharan Africa specifically
Ghanaian cities remains understudied. This is due to factors such as the
lack of air quality monitoring instruments, unenforced national emis-
sions standards, limited funding for research initiatives, insufficient
infrastructure for data collection and analysis, and disparities in the
prioritization of environmental health.

Notable among recent studies on ambient air pollution in Ghana
include a survey on the knowledge, perception and awareness of air
pollution by Odonkor and Mahami (2020). The study reported that most
of its respondents who were residents of the capital city, Accra, knew
about air pollution and its recurring impacts on their health. An air
quality study of 130 locations in the Greater Accra Metropolis by Wang
et al. (2024) predicted annual, harmattan and non-harmattan mean NO2
levels for the metropolis to be 37, 50, and 28 µg m− 3 respectively. The
study identified road traffic as the primary source of NO2 emissions with
the annual levels in the city exceeding the World Health Organization
(WHO) guideline of 10 µg m− 3. A PM2.5 monitoring study by Amegah
et al. (2022) in ten different traffic hotspots in Accra during the dry and
wet seasons reported median levels ranging from 27.08 – 51.37 and
20.04 – 48.92 µg m− 3 respectively. An increase of 1 µg m− 3 in PM2.5

exposure was associated with slight increases in respiratory and car-
diovascular, and symptoms among street traders. Nyadanu et al. (2022)
recorded an annual PM2.5 average of 59.97 µg m− 3 with an increase of
10 µg m− 3 in the annual average linked with a 3 % risk of stillbirth.

Existing studies on ambient air pollution in Ghana have focused on
the capital city, Accra, making the air quality status of other cities un-
known (Arku et al., 2008, 2015; Armah et al., 2010; Ofosu et al., 2012;
Rooney et al., 2012; Dionisio et al., 2010; Odonkor and Mahami 2020;
Kanhai et al., 2021; Amegah et al., 2022; Alli et al., 2023; Wang et al.,
2024). Expanding the evaluation of ambient air pollution beyond Accra
is crucial for a comprehensive assessment of air quality across Ghana.
Without up-to-date information on the trend of ambient air pollutants
for different urban cities in Ghana over the past two decades, policy-
makers will face challenges in identifying priority areas for intervention
and evaluating the efficacy of existing measures.

This study conducts a comprehensive health risk assessment of
ambient air pollutants in urban cities across Ghana from the year
2000–2023. It evaluates urban air quality and health risks of long-term
exposure to PM2.5, O3 and NO2 in the urban population of Ghana.
Furthermore, this study is crucial for addressing health inequalities as it
sheds light on how vulnerable populations such as the elderly and
children are affected by urban air quality. It provides essential infor-
mation to aid the tailoring of nation-specific public health interventions.
To the best of our knowledge, this study is the first to comprehensively
assess the ambient air quality of urban cities across different regions of
Ghana. This study seeks to inform policymakers, the scientific commu-
nity and public health officials about the status of urban air pollution in
Ghana. This study not only fills a critical gap on urban air quality in
Ghana but serves as a benchmark for facilitating informed decision-
making processes, equipping stakeholders with the knowledge needed
to develop effective evidence-based strategies to improve air quality and
safeguard the public health of urban populations in the country

2. Methodology

2.1. Study area

Ghana is situated in Sub-Saharan Africa, along the West African
coast, bordered by Côte d′Ivoire to the west, Burkina Faso to the north,
Togo to the east, and the Gulf of Guinea to the south. Its diverse land-
scape includes coastal plains, tropical rainforests, savannahs, and the
Volta River Basin. It covers a land area of approximately 240,000 km2

with a population of about 30 million people (Agodzo et al., 2023). A
majority of this population (57 %) reside in urban areas with almost half
of this urban population living in the Accra and Kumasi Metropolis
(Ghana Statistical Service, 2021; Iddrisu et al., 2023). The urban cities
involved in this study are illustrated in Fig. 1.

2.2. Data

This study used air quality observation data for 57 urban cities in
different regions of Ghana from 2000 to 2019. The data were obtained
from https://urbanairquality.online/ which provides a downloadable
dataset of air pollutant concentrations in all cities worldwide, as
described below. The obtained historical data from 2000 to 2019 for the
57 urban cities were used to estimate forecast values for 2020–2023 by
employing time series forecasting in Excel with a 95 % confidence in-
terval using the Exponential Smoothing (ETS) method. The retrieved
data included annual concentrations and population attributable frac-
tion (%) (PAF) for PM2.5, O3 and NO2 in 57 Ghanaian urban cities.

The obtained PM2.5 urban data were from a higher spatial resolution
dataset (1 km x 1 km) which integrated information from satellite-
retrieved aerosol optical depth, chemical transport modeling, and
ground monitor data incorporating geographically weighted regression.
Southerland et al. (2022) provide additional details on the method used
to generate the PM2.5 dataset. The obtained O3 data were from combined

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

https://urbanairquality.online/

ground measurement data with chemical transport model estimates,
downscaled to create finer resolution ozone concentration estimates
from 1990 to 2017 which were further extrapolated to 2019 based on
log-linear trends from 2008 to 2017 and re-gridded to a spatial resolu-
tion of 1 km to align with population estimates. Further information
regarding the method utilized to generate the O3 dataset can be found in
Malashock et al. (2022). The obtained NO2 data were from a resolution
of approximately 1 km2, calculated in 5-year increments between 1990
and 2010, and annually from 2010 to 2019. Additional details about the
methodology used in generating the NO2 dataset are provided by
Anenberg et al., (2022).

The concentrations in parts per billion (ppb) were converted to μg/
m3 using 1 ppb NO2 = 1.88 μg/m3 and 1 ppb O3 = 1.96 μg/m3 (Shen

et al., 2023). In addition, the population attributable fraction (%) (PAF)
for PM2.5, O3 and NO2 per annum also obtained from https://urbanai
rquality.online were used to estimate the mortality, disability-adjusted
life years (DALY), years of healthy life lost due to disability (YLD),
years of life lost from mortality (YLL) associated with ambient urban air
pollution in Ghana. These estimates were made for Chronic Respiratory
Diseases (CRD) and also specifically for Chronic obstructive pulmonary
disease (COPD), pneumoconiosis, asthma and Interstitial Lung disease
and pulmonary sarcoidosis (ILDS). The national burden of chronic res-
piratory diseases used for the estimates was retrieved from the Global
Burden of Disease Study in 2019 (Momtazmanesh et al., 2023).

Fig. 1. Map of Ghana showing the 57 urban cities in the study.

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

https://urbanairquality.online
https://urbanairquality.online

2.3. Ambient air quality

Ambient air quality is the extent to which the ambient air is free of
pollution and healthy for breathing. This can be assessed using a stan-
dard numerical rating known as the air quality index (AQI) (Table S1)
(Ababio et al., 2024). AQI is estimated from the concentrations of air
pollutants that are considered criteria for air quality. The sub-air quality
ratings of the ambient air pollutants (PM2.5, O3 and NO2) were used to
compute the air quality index. The worst sub-air quality indices were
used to adjudge the overall air quality index for each city and year in this
study. The sub-air quality indices for each air pollutant were calculated
using Eq. (1) (Abdul Raheem et al., 2022).

IP = (CP − BPLo)x
I Hi − I Lo

BP Hi − BP Lo
+ ILo (1)

where IP: Index value for the air pollutant
CP: Concentration of air pollutant
BP Lo: Lower Breakpoint value of CP
BP Hi: Higher Breakpoint value of CP
I Hi: Index Breakpoint value of BP Hi
I Lo: Index Breakpoint value of BP Lo
The breakpoint values for the concentrations of the air pollutants are

provided in Table S2. The six classifications of air quality index have
with their respective colour codes and ranges which are provided in
Table S1.

2.4. Dominant pollution types of PM2.5 and O3

The respective annual concentration thresholds [PM2.5 = 5 µg m− 3,
O3 = 60 µg m− 3] were used to categorize the dominant pollution types
of PM2.5 and O3 into four which are; Compound Pollution of PM2.5 and
O3 (P-O), PM2.5 Dominant Pollution, O3 Dominant Pollution, and Clean.
The categorization criteria which were modified from He et al. (2024)
are given in Table S3.

2.5. Analysis of synergistic changes in compound pollution

The relative rate of change (ROC) of PM2.5 and O3 concentrations in
2000 and 2023 was used to quantify the degree of synergistic changes in
pollution levels for the urban cities using Eq. 2 and Eq 3. (He et al.,
2024).

ROCi =
Ci, 2023

Ci, 2000
(2)

If =

⎧
⎪⎪⎨

⎪⎪⎩

ROCi,PM2.5 ≥1 and ROCi,O3 ≥1 Synergistic Increase
ROCi,PM2.5 <1 and ROCi,O3 <1 Synergistic Decrease

ROCi,PM2.5 ≥1 and ROCi,O3 <1 PM2.5 Increase and O3 Decrease
ROCi,PM2.5 <1 and ROCi,O3 ≤1 PM2.5 Decrease and O3 Increase

(3)

2.6. Health risk assessment

2.6.1. Daily dose
Daily Dose (DD) is the chronic inhalation of an ambient air pollutant

daily. DD was calculated using Eq. (4) which was modified from (Ababio
et al., 2023). The parameter values for different age groups used for
evaluating the corresponding health risks for 1 h, 3 h, 6 h, 9 h and 12 h
of exposure to the air pollutants have been provided in Table 1.

DD =
C x R x T x F x D

BWxAT
(4)

where C: Concentration of PM2.5 or NO2 or O3
R: rate of inhalation (m3 hr− 1)
T: time of exposure
F: frequency of exposure (365 d yr− 1)

D: duration of exposure
BW: average body weight
AT: average time (duration of exposure x frequency of exposure)
The parameter values used for estimating the daily dose in the study

were adopted from the exposure factor handbook of the United States
Environmental Protection Agency (2011). The employed values are
provided in Table 1.

2.6.2. Non-carcinogenic risk
As a ratio between the daily dose and reference dose, the Hazard

Quotient (HQ) reveals how the ambient air pollutant compares to its
reference dose. The hazard quotient and Hazard index (HI) are in-
dicators of non-carcinogenic health risks calculated using Eq. 5 and Eq.
6 (Ababio et al., 2023). HI predicts the sum of the individual HQ for the
criteria pollutants through an exposure route to the reference dose (RfD)
for each pollutant (Hogarh et al., 2018). An HI/HQ < 1 and HI/HQ > 1
imply negligible non-cancer health risk and a higher chronic non-cancer
risk respectively.

HQ =
DD
RfD

(5)

HI = HQPM2.5 + HQNO2 + HQO3 (6)

The reference dose (RfD) determined using Eq. (7) is the concen-
tration at which daily exposure to a pollutant such as PM2.5, O3, and NO2
does not cause adverse health effects over a lifetime (Ababio et al.,
2023).

RfD =
RfC x IR

BW
(7)

Where RfC: Reference concentration (WHO annual guideline values
PM2.5 = 5 µg m− 3, O3 = 60 µg m− 3 and NO2 = 10 µg m− 3 (World Health
Organization, 2021)

IR: Inhalation rate (m3 d− 1)
BW: average body weight

2.6.3. Carcinogenic risk for PM2.5 exposure
The carcinogenic risk which is the incremental chance of developing

cancer from exposure to ambient PM2.5 in urban cities was calculated
using Eq. (8). The accepted range for carcinogenic risk is 10− 4 to 10− 6,
as risks < 10− 6 are considered negligible and > 10− 4 are considered
highly unacceptable and a cause for alarm.

CR = DD x SF (8)

The slope factor (SF) was used to estimate the likelihood of devel-
oping cancer from exposure to a carcinogenic pollutant such as ambient
PM2.5. SF was calculated using Eq. (9) for adults, adolescents, children,
toddlers and infants (Mbazima, 2022).

SF =
IUR

BWxIR
(9)

Where SF: Slope Factor
IUR: Inhalation Unit Risk (0.008)
IR: Inhalation rate
BW: average body weight

Table 1
Parameter values for estimating the daily dose in different age groups.

Age Group
Inhalation rate (m3

d¡1)
Exposure duration
(yrs)

Body weight
(kg)

Infant 5 1 7
Toddler 9 4 12
Children 12 12 23
Adolescents 16 19 62
Adults 20 70 70

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

2.6.4. Relative and excess risk

2.6.4.1. Relative risk. Relative risk (RR) is a measure of the degree of
correlation between exposure to a particular pollutant and the likeli-
hood of developing health outcomes such as a disease or mortality. RR
was obtained using Eq. (10) (Sharma et al., 2023).

RR = exp [β (C − Co)] (10)

Where β: the exposure-response coefficient is the corresponding
additional health risk per unit increase in an air pollutant beyond its
threshold concentration. β = 0.038 %, 0.13 % and 0.048 % per 1 µg m− 3

increase of PM2.5, NO2 and O3 respectively. (Shen et al., 2020)
C: Concentration of PM2.5 or NO2 or O3
Co: the threshold concentration (WHO annual guideline values)

2.6.4.2. Excess risk. Excess Risk (ER) is the difference in risk between
exposed and unexposed individuals. It was used to estimate the addi-
tional risk of health outcomes or mortalities due to exposure to the
pollutant. ER was computed using Eq. (11) (Sharma et al., 2023)

ER = RR – 1 (11)

Where RR: Relative Risk

2.6.5. Air quality life index
The Air Quality Life Index (AQLI) estimates the impact of air pollu-

tion on life expectancy (Greenstone and Fan, 2018). It measures the
potential gains in life expectancy per person (Life Years Saved) if PM2.5
concentrations meet WHO or national air quality standards. It is based
on findings that sustained exposure to an additional 10 µg m− 3 of PM2.5
above the recommended threshold reduces life expectancy by 0.98 years
(Chen et al., 2013; Ebenstein et al., 2017; Greenstone and Fan, 2018;
Rana, 2022). It was calculated using Eq. (12)

Life Years Saved =

(
C − S

)

x 0.98 (12)

Where C = the annual concentration of PM2.5 in µg m− 3

S = PM2.5 WHO annual guideline value = 5 µg m− 3

2.6.6. Mortality, Disability-Adjusted Life Years (DALY), Years of Healthy
Life Lost due to Disability (YLD) and Years of Life Lost from Mortality (YLL)
estimates for all age groups

The estimates for the deaths, DALY, YLD and YLL associated with
ambient concentrations of PM2.5, NO2 and O3 were made to assess the
impact on respiratory health burden in Ghana. Using the annual median
population attributable fraction (%) (PAF) data spanning from 2000 to
2023, the adjusted deaths, DALYs, YLLs, and YLDs for CRD and specific
respiratory diseases (COPD, pneumoconiosis, asthma, and ILSD) were
calculated accordingly using Eqs. (13), (14), (15), (16), (17) and (18).
The mortality, DALY, YLL and YLD values for Ghana from the Global
Burden of Disease Study in 2019 (Momtazmanesh et al., 2023) used for
the estimation are provided in Table S4.

IncidencePM2.5 or NO2 or O3 = Mortalitydisease x PAFPM2.5 or NO2 or O3 (13)

PrevalencePM2.5 or NO2 or O3 = Mortalitydisease x PAFPM2.5 or NO2 or O3 (14)

MortalityPM2.5 or NO2 or O3
= Mortalitydisease x PAFPM2.5 or NO2 or O3 (15)

DALYPM2.5 or NO2 or O3 = Mortalitydisease x PAFPM2.5 or NO2 or O3 (16)

YLLPM2.5 or NO2 or O3 = Mortalitydisease x PAFPM2.5 or NO2 or O3 (17)

YLDPM2.5 or NO2 or O3 = Mortalitydisease x PAFPM2.5 or NO2 or O3 (18)

2.7. Statistical analysis

The retrieved data underwent statistical analysis using the Stata and
Microsoft Excel. Quantile regression was employed in this study to
analyze the relationship between ambient air pollution in urban cities.

It was used in this study due to the skewed distribution exhibited by
the obtained data. It was used to estimate the median and other quan-
tiles of ambient air pollutant concentrations. It allowed for the com-
parison of the quantile estimates across urban cities. It made it possible
to assess the differences in pollutant concentrations for the different
years. Additionally, regression coefficients (β) were also estimated from
quantile regressions.

The p-values associated with the regression coefficients were also
generated using quantile regression which indicated the statistical sig-
nificance of the differences in pollutant levels between the cities.
Moreover, the Kendall Tau (KT) test was also used to assess trends in the
time series data of the air pollutants.

The Kendall’s tau coefficients (τₐ and τᵦ) and the associated p-value
were estimated. Kendall’s tau coefficients were used to measure the
association between the variables "year" and "pollutant concentration",
with positive values indicating an increasing trend, negative values
indicating a decreasing trend, and values close to zero indicating no
significant trend. The associated p-value provided a measure of the
statistical significance of the observed trend. In addition, Sen’s slope was
also employed to analyze trends in pollutant concentrations over time
using the estimated slope coefficient, p-value which determined the
significance level, and the confidence interval which assessed the pre-
cision and reliability of the estimated trend. Moreover, Principal
Component Analysis (PCA) was used to analyze air pollutant patterns in
the various urban cities compressing the dataset into components to
reveal the relationship between pollutants and their spatial distribution.

3. Results and discussion

3.1. Ambient air pollutant concentrations, population attributable
fraction (%), epidemiological impacts and air quality indices of the urban
cities

The annual concentrations of PM2.5, NO2, and O3 for different urban
cities in Ghana from 2000 to 2023 are reported as median and quartiles
in Table 2. The corresponding p-values, regression coefficients, and
confidence intervals for the annual concentrations of PM2.5, NO2, and O3
using the capital city, Accra as the reference are provided in Tables S1
and S2. The annual PM2.5, and O3 concentrations for each city in the
study are provided in Fig. S1a – Fig. S1e and that of NO2 in Fig. S2a –
Fig. S2e.

The annual median PM2.5 concentrations of the urban cities spanned
from 50.79 to 67.97 µg m− 3. These were five to six-fold higher than the
WHO annual PM2.5 air quality guideline value (5 µg m− 3). Spatial dis-
tribution revealed the urban populace in Akatsi and Jirapa situated in
Southeastern and Northwestern Ghana to be exposed to the highest and
lowest median annual PM2.5 concentrations respectively (Fig. 2a).

It was observed that cities such as Accra, Koforidua, and Kumasi in
the Southern regions of Ghana consistently exhibited higher annual
median PM2.5 concentrations compared to lower levels in cities in the
Northern belt such as Wa, Tamale, and Nalerigu. The PM2.5 findings of
this study were similar to the findings of Moro et al. (2022) as Savannah
belt cities had the lowest PM2.5 concentrations, followed by Forest belt
and Coastal belt cities. The inter-urban comparison of PM2.5 concen-
trations using the capital city, Accra, as the reference indicated 20 of the
urban cities in the study to have positive regression coefficients (β) (0.31
– 7.80) (Table S5). These revealed residents of these cities to be exposed
to higher levels of PM2.5 as compared to urban dwellers in the capital.
These might be attributed to anthropogenic factors such as higher reli-
ance on traditional biomass cooking sources, lack of green spaces,
untarred roads, and combustion of waste coupled with meteorological

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

conditions such as higher temperatures and low wind speeds that could
have exacerbated PM2.5 pollution in these cities. Negative regression
coefficients (β) (-0.40 to − 10.40) revealed the majority (36) of residents
in the urban cities in this study to be exposed to lower levels of PM2.5 as
compared to the capital. These might have been due to lower population
densities, limited industrial activities and vehicular traffic. Furthermore,
fine particulate matter (PM2.5) concentrations in the urban cities except
Jirapa and Wa were not significantly different from the capital with
p > 0.05 (Table S5).

The AQLI estimates for gains in life expectancy if ambient PM2.5
concentrations per annum in the urban cities met WHO guidelines are

given in Fig. 3 and Table S6. The AQLI revealed that ambient PM2.5
pollution in urban cities cut life expectancy short by 4.5–6.2 years
relative to if concentrations in the cities met the WHO threshold. This
revealed the detrimental impact of PM2.5 pollution on the life expec-
tancy of habitants in Ghanaian urban cities.

The relative rate of changes for P-O compound pollution and the
dominant pollution classifications from the ambient concentrations of
PM2.5 and O3 in the urban cities are provided in Table S7. The ambient
environment of all the urban cities in the study except Cape Coast and
Sekondi Takoradi exhibited a PM2.5-O3 dominant compound pollution
as concentrations of PM2.5 > 5 µg m− 3 and O3 > 31 ppb. In the absence
of corresponding ozone concentrations, the ambient environs of Cape
Coast and Sekondi-Takoradi were classified as PM2.5 dominated pollu-
tion as PM2.5 > 5 µg m− 3. The relative rate of changes revealed PM2.5-O3
compound pollution in Sekondi-Takoradi and Cape Coast to have syn-
ergistic changes of PM2.5 decrease. The observed level of synergistic
changes for ambient PM2.5-O3 pollution in Akim Oda and Mampong was
synergistic decreases in P-O compound pollution. In contrast, the syn-
ergistic increase in P-O compound pollution was observed for Nkawkaw
for the study’s duration. The remaining 54 urban cities exhibited syn-
ergistic changes of PM2.5 decrease and O3 increase for P-O compound
pollution from 2000 to 2023. The findings for urban ambient environ-
ments in this study was similar to a comparative study by He et al.
(2024) which observed P-O compound pollution in majority of global
cities, predominantly in urban Asian cities.

The median tropospheric ozone concentrations in the urban cities
varied from 72.21 – 92.58 µg m− 3 which exceeded the WHO annual
guideline equivalent recommendation of 60 µg m− 3. The highest and
lowest median concentrations of ozone were found in Tatale and the
capital, Accra, situated in the Northeastern and Southern parts of the
country respectively (Fig. 2b).

Inter-urban comparison indicated cities in Southern Ghana had
lower median concentrations of ground-level ozone as compared to
higher concentrations in Northern Ghana cities. Furthermore, a com-
parison of the ozone concentrations in reference to the capital city only
gave positive regression coefficients (β) (0.36 – 22.26) (Table S5). This
suggested that the urban population of the other cities in the study were
exposed to higher concentrations of ground-level ozone than in the
capital. This could have been due to factors like cleaner energy initia-
tives and favourable meteorological conditions for dispersion and
degradation that led to the lesser emission of ozone precursors such as
nitrogen oxides (NOx) and volatile organic compounds (VOCs) in Accra.
Tropospheric ozone concentrations in 33 urban cities were significantly
different from that of the capital with p < 0.05 while those of 21 cities
were not with p > 0.05 (Table S5).

Median NO2 concentrations of the urban cities ranged from 3.65 –
12.15 µg m− 3. The highest and lowest were observed in Garu and Cape
Coast accordingly. Few of the urban cities had nitrogen dioxide con-
centrations exceeding the WHO guideline value of 10 µg m− 3. Urban
cities situated within the Upper East and Northern Regions of Ghana
recorded the highest NO2 concentrations whiles cities in the Central
Region of Ghana recorded the lowest (Fig. 2.c).

Positive regression coefficients (β) (0.30 – 0.94) revealed the resi-
dents of Bawku, Garu. and Nalerigu to be exposed to concentrations of
NO2 above that of the capital. Furthermore p > 0.05 indicated the NO2
concentrations in Bolgatanga, Bawku, and Wa to be significantly
different from that of the capital.

The median PM2.5 concentrations per annum varied from 46.69 –
82.52 µg m− 3 (Fig. 4A). These were 5 – 8 times above the WHO rec-
ommended threshold. The highest was observed in 2000 while the
lowest was observed in 2014. The PM2.5 concentrations exhibited fluc-
tuations across the 23-year duration. The years 2000 – 2001, 2002 –
2003, 2004 – 2006, 2008 – 2009, and 2012 – 2013 exhibited significant
decreases in PM2.5 concentrations while significant increases were
observed for the years, 2001 – 2002, 2003 – 2004, 2006 – 2007, 2014 –
2015, and 2019 – 2020. In comparison with the year 2000, negative

Table 2
Concentrations of PM2.5, NO2 and O3 in Ghanaian Urban Cities.

City
PM2.5 (µg m¡3) O3 (µg m¡3) NO2 (µg m¡3)

Median IQR Median IQR Median IQR

Accra 61.43 13.53 72.21 24.09 11.06 0.95
Agona Swedru 62.98 15.93 73.04 24.41 6.25 0.48
Akatsi 67.97 11.98 78.48 24.99 6.96 0.53
Akim Oda 62.33 19.08 80.09 24.34 6.65 0.34
Asamankese 63.42 16.18 80.59 24.33 4.94 0.35
Assin Fosu 64.80 20.33 74.81 23.30 6.14 0.67
Atebubu 61.93 11.77 89.37 22.98 5.97 1.18
Bawku 58.22 12.47 85.94 16.93 11.43 0.92
Berekum 55.72 13.67 83.99 22.94 7.87 0.76
Bimbila 61.17 9.11 92.01 19.06 8.41 0.93
Bole 56.17 8.90 88.79 15.13 7.91 2.79
Bolgatanga 56.34 9.31 86.43 17.19 11.08 0.95
Cape Coast 60.27 16.95 nd Nd 3.65 0.28
Dambai 64.56 10.96 90.15 21.42 7.34 1.26
Dormaa Ahenkro 57.64 17.35 80.81 22.96 6.82 0.95
Dunkwa Offin 60.55 17.83 78.23 22.45 5.50 0.57
Ejura 60.29 11.11 87.62 23.87 7.27 0.87
Enchi 59.43 16.26 72.99 21.65 4.69 0.48
Garu 58.07 13.52 87.47 17.55 12.15 0.84
Gushiegu 56.28 8.86 90.88 18.06 10.01 0.70
Ho 63.58 8.00 83.50 26.12 5.88 2.02
Hohoe 64.24 11.20 87.52 24.84 5.31 0.50
Jirapa 50.79 6.87 85.42 16.42 9.08 1.42
Karaga 54.49 10.73 90.59 18.06 9.76 1.84
Kete Krachi 61.91 10.80 90.08 23.62 6.54 1.02
Kintampo 58.24 12.90 88.10 19.60 5.94 1.05
Koforidua 61.14 13.31 81.10 26.05 7.05 0.83
Konongo 59.93 15.82 82.33 24.56 7.00 1.08
Kumasi 63.15 18.65 83.42 24.24 10.27 0.32
Kwame Danso 62.22 12.73 88.77 23.09 6.16 2.61
Mampong 58.81 11.31 85.81 23.88 6.27 0.61
Nalerigu 54.80 7.66 87.49 17.55 11.58 0.40
Navrongo 56.48 9.86 85.65 16.83 9.82 0.67
New Tafo Akim 60.03 13.66 82.55 26.03 6.46 0.45
Nkawkaw 60.43 12.79 83.48 25.47 6.49 0.71
Nkoranza 57.76 12.81 86.33 21.96 5.84 0.86
Nsawam 62.34 15.32 82.35 24.12 8.05 0.47
Obuasi 61.43 18.83 77.51 23.89 7.48 0.56
Paga 55.52 9.27 85.09 16.72 8.86 0.71
Salaga 62.03 9.01 89.40 19.41 7.15 1.78
Sandema 55.83 11.15 85.89 16.83 8.89 1.83
Savelugu 55.60 9.74 90.33 17.81 9.08 1.51
Sawla 55.93 9.48 89.16 15.32 7.52 2.64
Somanya 60.51 11.03 81.45 25.73 6.07 0.60
Suhum 62.09 16.81 80.42 26.06 5.61 0.52
Sunyani 57.26 13.10 83.95 23.36 9.11 1.47
Sekondi-Takoradi 59.27 15.04 nd Nd 4.66 0.25
Tamale 57.21 11.88 91.43 18.28 10.45 1.42
Tatale 57.71 9.92 92.58 17.45 8.21 1.51
Techiman 57.18 12.55 86.31 21.96 7.35 0.42
Wa 51.57 8.70 86.78 15.87 10.62 2.19
Walewale 55.34 9.80 88.27 17.58 10.17 1.14
Winneba 63.73 16.18 73.58 22.28 5.83 0.41
Wulensi 63.58 10.28 91.23 19.21 7.97 1.67
Yeji 61.65 9.41 89.86 21.19 7.29 1.82
Yendi 55.77 11.38 92.21 18.00 9.72 1.46
Zebilla 56.50 10.92 86.51 17.50 9.80 0.68

IQR means Interquartile range; nd means no data

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

quantile regression coefficients (β) (-4.90 to − 35.83) showed significant
reductions in PM2.5 concentrations in the urban cities per annum
(Table S6). The highest and lowest reductions in annual PM2.5 concen-
trations were observed in the years 2000 and 2014 respectively. The
median PM2.5 concentrations in the urban cities per annum were
significantly different as p < 0.001 (Table S8). Negative Kendall tau
coefficients (τ) (τₐ= − 0.1354 and τᵦ = − 0.1383) and a p-value < 0.0001
were obtained from the KT test which revealed a weak significant
negative correlation between years and PM2.5 concentrations (Table S9).

Also, negative slope coefficient β (-0.14), a p-value < 0.001 and 95 % CI
[-0.17, − 0.11] were obtained from the Sen’s slope analysis (Table S10).
These indicated a significant decreasing trend in PM2.5 concentrations
over the observed years. This could have been due to the transition from
biomass combustion sources to cleaner energies such as the increased
adoption of liquified petroleum gas for cooking in urban households.
Increased greening initiatives such as tree planting programs and the
tarring of unpaved roads and streets in urban municipalities reducing
the resuspension of dust. Moreover, stringent air quality policies

Fig. 2. (a). Spatial Distribution of PM2.5 Concentrations in Urban Cities in Ghana (b) Spatial Distribution of O3 Concentrations in Urban Cities in Ghana (c) Spatial
Distribution of NO2 Concentrations in Urban Cities in Ghana.

Fig. 3. Life Years Saved per Annum if PM2.5 Concentrations met Recommended Threshold.

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

curtailing particulate emissions from industrial activities, together with
improved urban waste management reducing the open burning of waste
might have accounted for the observed PM2.5 decline. Median ground-
level ozone concentrations per year ranged from 64.36 –
105.37 µg m− 3 which were in exceedance of the WHO guideline limit
(Fig. 2A). The highest and lowest were recorded for the years 2021 and
2003 respectively. A trend of fluctuations and plateaus was observed for
ozone concentrations across the 23-year duration. Noticeable plateaus in
ozone concentrations were observed for the years, 2000 – 2001, 2006 –
2007, 2016 – 2017 and 2020 – 2021. Relatively rapid increases in ozone
occurred from 2004 – 2006, 2009 – 2015. Notable decreases were
observed from 2001 – 2004, 2007 – 2009 and 2015 – 2016. Further-
more, in the last few years (2021 – 2023), there has been a significant
decrease in ozone levels following the peak in 2021. The variability in
ground-level ozone levels may have been due to the complex interplay of
factors such as fluctuating weather conditions, including temperature,
humidity, and wind speed, which might have influenced the formation
and dispersion of ozone. Also, anthropogenic activities such as industrial
processes and transportation emissions might have contributed to ozone
formation through the emission of precursor pollutants. The KT test

indicated a strong positive correlation between years and O3 concen-
tration, with Kendall’s tau coefficients (τ) of 0.6097 (τₐ) and 0.6229 (τᵦ).
In addition, a p-value < 0.0001 indicated a significant increasing trend
in tropospheric O3 concentration over the 23 years (Table S9). In
addition, a positive slope coefficient β (0.61), a p-value < 0.001 and
95 % CI [0.59, 0.63] were obtained from the Sen’s slope analysis
(Table S10). These confirmed a consistently increasing trend in O3
concentrations over the years in urban cities. This could have been due
to the continuous decline in Nitrogen Oxides (NOx) accounting for an
increase in the ratio of Volatile Organic Compounds (VOCs) to NOx. This
might have favoured ozone formation in a NOx-limited regime through
the suppression of ozone titration, increased radical activities and
peroxide formation over nitric acid formation. In addition, higher tem-
peratures increasing ozone production by rising emissions of precursors
such as VOCs alongside higher solar radiation intensities in the study
area might have accounted for the observed increase in tropospheric
ozone.

Median NO2 concentrations per year in the urban cities spanned from
6.41 – 8.18 µg m− 3 below the WHO threshold (Fig. 4B). The lowest and
highest concentrations were recorded in the years 2023 and 2005

Fig. 4. (a) Annual Median Concentrations of PM2.5 and O3 in Ghanaian Urban Cities (b) Annual Median Concentrations of NO2 in Ghanaian Urban Cities.

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

respectively. A general trend of decreasing NO2 concentrations was
observed with variabilities in NO2 concentrations from year to year as
some years showed significant increases (2000 – 2005, 2010 – 2011,
2014 – 2015) or decreases (2005 – 2010, 2017 – 2019) compared to the
preceding years, indicating fluctuations in NO2 levels. The general
declining trend in urban NO2 concentrations appears to have stabilized
in the most recent years (2021–2023), indicating that the trend may
have slowed down or plateaued. Kendall’s tau coefficients (τ) [-0.1192
(τₐ) and − 0.1218 (τᵦ)] and a p-value < 0.0001 indicated a weak nega-
tive correlation between years and NO2 concentrations (Table S9).
Furthermore, a negative slope coefficient β (-0.12), a p-value < 0.001
and 95 % CI [-0.15, − 0.08] were obtained from the Sen’s slope analysis
(Table S10). These suggested a significant decreasing trend in NO2
concentration over the years. This could be attributed to improved
regulations of vehicular emissions and industrial processes through the
adoption of cleaner energy alternatives, usage of catalytic converters in
vehicles, improved public transit systems coupled with meteorological
conditions such as dispersing wind patterns might have led to the
decreasing trend in ambient urban NO2 concentrations.

Comparing the annual concentrations of PM2.5 in Ghanaian urban
cities in this study to those in other global regions revealed notable
differences (Table 3). Research by Tariq and colleagues (2023), and
Mandal et al. (2020) reported higher concentrations in Niger and India

respectively. The findings of Liu et al. (2023), Xu and Zhang (2020) and
Rahman et al. (2021) reported a broader range of annual concentrations
in urban Asian cities. In contrast, Park et al. (2020) reported a smaller
range of annual concentrations in Seoul. Furthermore, Li et al. (2021)
observed lower concentrations in Iraq and Kuwait. Azhari et al. (2021)
also reported lower concentration ranges in Malaysia. Moreover, a local
study by Bahino et al. (2024) reported notably lower concentrations in
Accra, Ghana than this study.

The comparison of tropospheric ozone concentrations in the study
with other studies in different geographical locations also revealed sig-
nificant variability. The research conducted by He et al. (2023) reported
wider concentration ranges compared to this study. The reported annual
ranges in this study were above the reported levels of McHugh et al.
(2023) and Xu et al. (2019). Higher and ground O3 concentration ranges
than this study’s findings were observed by Huang et al. (2018) and Lee
et al. (2023) in China. The annual nitrogen dioxide findings of this study
were lower than the reported concentrations from related studies in
France and China (Jurado et al., 2020; Sanyal et al., 2018; Kuerban
et al., 2020). The differences in concentrations observed in this study
could be due to its longer duration compared to related studies such as
Wang et al. (2024) which had a shorter timeframe.

Principal Component Analysis (PCA) was employed in this study to
determine spatial variability of air pollutants and the influence of each
pollutant on principal components using three-factor loading categories;
strong factors (> 0.75), moderate (0.50 – 0.75), and weak factors (<
0.50) (Ashong et al., 2024). The PCA revealed three principal compo-
nents (total variance); PC1 (62.02 %), PC2 (22.72 %) and PC3 (15.26 %)
in the study (Fig. 5). The first principal component (PC1) represented
the most dominant air pollution pattern in the urban cities revealing an
inverse relationship between PM2.5 (- 0.61) and the gaseous pollutants;
NO2 (0.60) and O3 (0.52). This could have been due to the prevalence of
particulate emission sources (such as the combustion of solid fuels and
dust from construction and mining activities) over gaseous pollutant
sources. Also, finer PM2.5 providing surfaces for heterogeneous reactions
might have favoured the scavenging of gaseous pollutants in the atmo-
sphere. The second principal component (PC2) characterized by a strong
positive loading of O3 (0.85) and a negative loading of NO2 (- 0.45)
indicated an inverse relationship between the two gaseous pollutants.
This could be due to a non-linear photochemical ozone formation with
NO2 as a precursor in urban environs where higher solar radiation en-
hances ozone formation with lower NO2 reducing ozone titration. This
suggests PC2 represents photochemical processes governing ozone for-
mation and depletion in the troposphere. The third principal component
(PC3) showed strong positive loadings for both PM2.5 (0.74) and NO2
(0.67) with minimal influence from O3 (0.1). This component likely
represents localized pollution patterns and specific pollution sources
affecting both PM2.5 and NO2 concentrations simultaneously. The
co-variation of PM2.5 and NO2 in PC3 could be due to common pollution
sources such as high traffic areas and industrial zones in urban cities.

The dominant role of PC1 suggests that air pollution mitigation
strategies targeting PM2.5 reduction may not proportionally affect NO2
and O3. Necessitating multi-pollutant approaches to addressing urban
air pollution in Ghana. Furthermore, the distinct behaviour of O3 in PC2
highlights the need for interventions tailored to target ozone formation
particularly in urban areas prone to photochemical smog. In addition,
PC3 suggests the possibility of some targeted interventions mitigating
both PM2.5 and NO2 simultaneously in certain urban areas. The PCA
biplot revealed southern urban cities exhibited elevated concentrations
of PM2.which could be attributed to higher vehicular traffic, industrial
processes and combustion activities in southern Ghana. In contrast,
northern urban cities had higher O3 concentrations likely due to
photochemical reactions involving NO2 and VOC precursors under
favourable conditions such as higher solar radiations in northern Ghana.

The population attributable fraction (%) (PAF) for air pollution is an
epidemiologic measure which estimates the proportion of disease cases
in a population that can be attributed to exposure to a specific air

Table 3
Comparison of concentrations of ambient urban air pollutants with comparative
studies.

Reference Location Pollutant Study
Duration

Annual
Concentrations

This Study
Ghana

PM2.5

2000 –
2023

46.69 –
82.52 µg m− 3

Tariq et al.
(2023) Niger

1998 –
2019

68.85 –
70.47 µg m− 3

Liu et al.
(2023)

China 2013 –
2021

6.5 – 94.2 µg m− 3

Xu and Zhang
(2020)

Beijing, China 2013 –
2018

5 – 470 µg m− 3

Park et al.
(2020)

Seoul, South
Korea

2014 –
2015 41 – 46.7 µg m− 3

Mandal et al.
(2020) Delhi, India

2010 –
2016 87 – 138 µg m− 3

Li et al.
(2021)

Iraq 2001 –
2018

33 – 44 µg m− 3

Kuwait 36 – 49 µg m− 3

Azhari et al.
(2021)

Kuala Lumpur,
Malaysia

2019 30.4 – 43.7 µg m− 3

Rahman et al.
(2021)

Dhaka,
Bangladesh 2016 5.27 – 105 µg m− 3

Bahino et al.
(2024) Accra, Ghana

2020 –
2021 21.1 – 24.8 µg m− 3

This Study Ghana

2000 –
2023

64.36 –
105.37 µg m− 3

He et al.
(2023)

China 2013 –
2018

1.84 – 160 µg m− 3

McHugh
et al. (2023) Ireland

2010 –
2019 39 – 43 µg m− 3

Lee et al.
(2023)

California,
USA

2005 –
2017

94.08 –
103.88 µg m− 3

Huang et al.
(2018)

China 2013 –
2017

141.0 –
163.5 µg m− 3

Xu et al.
(2019)

Windsor,
Canada

1996 –
2015

39.79 –
52.92 µg m− 3

This Study
Ghana

NO2

2000 –
2023 6.41 – 8.18 µg m− 3

Jurado et al.
(2020)

France
2013 –
2017

5 – 95 µg m− 3

Kuerban
et al. (2020)

China 2015 –
2018

5.9–64.4 µg m− 3

Sanyal et al.
(2018) France

1999 –
2000

4.55 –
46.96 µg m− 3

Wang et al.
(2022) Accra, Ghana

2019 –
2020 70 µg m− 3

Wang et al.
(2024)

Accra, Ghana
2019 –
2020

37 µg m− 3

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

pollutant. Thus, it represents the percentage of disease burden in the
population that could be avoided if exposure to the air pollutant were
mitigated. The median PAF for annual air pollutant exposure in the
urban cities is given in Table S11 and S12.The median PAF associated
with ambient exposure to PM2.5, O3, and NO2 in the urban cities varied
from 14.25 – 34.61 %, 1.94 – 8.28 %, and 1.14 – 11.72 % respectively.
The highest annual PAF for PM2.5, O3, and NO2 exposure were observed
for the years 2000, 2020 and 2005 respectively (Table S12). The Mann-
Kendall test was used to examine the relationship between years and
PAFs for PM2.5, O3, and NO2. The obtained Kendall’s tau coefficients (τ)
[PM2.5PAF = -0.1326 (τₐ) and − 0.1366 (τᵦ); NO2PAF = -0.0804 (τₐ) and
− 0.0821 (τᵦ)] revealed weak negative correlations for PM2.5PAF and
NO2PAF with years (Table S12). Additionally, p-values < 0.0001
showed a significant decreasing trend in PM2.5PAF and NO2PAF over the
years (Table S13). This suggested existing interventions in Ghana to
tackle PM2.5PAF and NO2PAF were effective. In contrast, positive Ken-
dall’s tau coefficients (τ) [-0.1326 (τₐ) and − 0.1366 (τᵦ)] a p-value
< 0.0001 were obtained for O3PAF which indicated a significant
increasing trend over the years. This indicated the need for targeted
strategies to address O3PAF in Ghana.

The epidemiological contribution of PM2.5, NO2, and O3 to in-
cidences, prevalences, mortalities, DALYs, YLLs and YLDs in Ghana were
estimated for chronic respiratory diseases in general and specifically for
COPD, pneumoconiosis, asthma, and ILSD. PM2.5 had the highest inci-
dence of respiratory diseases. CRD had the most cases at 63190, fol-
lowed by asthma with 43296 cases. COPD and ILDS also recorded high
incidences of 7019 and 12847 cases, respectively. PCN had the lowest
incidence among the diseases, with only 28 cases attributed to PM2.5
exposure. Exposure to ground-level ozone also resulted in the incidence
of respiratory diseases albeit with a reduced impact compared to PM2.5.
CRD had the highest incidence (12667) followed by asthma (8679).

COPD and ILDS had lower incidences of 1407 and 2575 cases, respec-
tively. PCN registered the lowest incidence, with only 6 cases linked to
ambient O3 exposure. Nitrogen dioxide had the lowest incidence among
the three air pollutants as CRD had the highest incidence at 9239 cases
followed by asthma with 6331 cases. COPD and ILDS had lower in-
cidences of 1026 and 1878 cases, respectively. PCN has the lowest
incidence, with only 4 cases attributed to NO2 exposure (Table S14). The
estimated prevalence of CRD for annual PM2.5 exposure was found to be
329880, which is significantly higher than the prevalence of COPD
(115965), PCN (59), asthma (245644), and ILDS (1736). The prevalence
of CRD due to O3 exposure was estimated to be 66127, while that for
NO2 is 48233. The prevalence from ambient urban O3 had estimates of
23246 for COPD, 12 for PCN, 43950 for asthma, and 348 for ILDS. NO2
had the lowest prevalence among the three pollutants with prevalence of
16956 for COPD, 9 for PCN, 32057 for asthma, and 254 for ILDS
(Table S15).

The estimated annual CRD mortalities varied from 1734 – 2185, 45 –
704, and 233 – 344 for ambient exposure to PM2.5, O3, and NO2
respectively. The average annual estimated deaths for COPD, pneumo-
coniosis, asthma, and ILSD from ambient PM2.5 were 1079, 1, 696 and
42 respectively. Tropospheric ozone was estimated to account for 391,
216, 140 and 8 annual average mortalities in Ghana from CRD, COPD,
asthma, and ILSD respectively. Furthermore, the ambient NO2 in the
study was found to result in 285, 158, 102 and 6 annual average mor-
talities in Ghana from CRD, COPD, asthma, and ILSD respectively. Es-
timates from the study indicated no mortalities for pneumoconiosis from
O3 and NO2 ambient urban exposure in Ghana (Table S16). The annual
DALY estimates revealed substantial impacts of the urban air pollutants
on CRD, COPD, PCN, Asthma, and ILDS. It was observed that PM2.5
consistently had the highest annual DALY average, generally for CRD
(78653) and particularly for COPD (37933). These showed it

Fig. 5. Principal component analysis biplot of air pollutants in Ghanaian urban cities.

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

significantly exacerbated CRDs in comparison with the other pollutants
in Ghana. Annual average DALYs of 11500 (NO2) and 15767 (O3) were
observed for CRD. In addition, DALYs linked to COPs were 7604 and
5546 for O3 and NO2 respectively (Table S17).

The obtained annual average YLL from CRD, COPD, PCN asthma, and
ILSD ranged from 34 – 53938, 7 – 10812, 5 – 7887, for PM2.5, O3, and
NO2 ambient exposure respectively. The highest and lowest YLLs were
observed for CRD and PCN respectively. The highest YLLs for all the
diseases were consistently associated with ambient PM2.5 in the urban
cities (Table S18). The highest and lowest annual YLD averages were
observed for CRD (24715) and particularly for COPD (13456) respec-
tively. The YLD estimates from the study indicated PM2.5 had the most
substantial impact on YLD, with notably high YLDs for CPD (24715),
COPD (13456), asthma (8664) and ILDS (197). In comparison, O3
recorded lower YLDs for CRD (4954), COPD (2697), asthma (1737), and
ILDS (40). On the other hand, NO2 registered the lowest YLDs for CRD
(3614), COPD (1967), asthma (1267) and ILDS (29) among the three
pollutants. The YLD estimates for pneumoconiosis associated with
exposure to the three pollutants were < 10 (Table S19).

The epidemiological contribution of PM2.5, O3, and NO2 to CRD in
Ghana revealed a significant public health burden with PM2.5 consis-
tently showing the highest impact across all health outcomes. The clear
hierarchy of impact for incidence, prevalence, mortality, DALY, YLD and
YLL in Ghana was PM2.5 > O3 > NO2. Notably, CRD bore the highest
brunt of air pollution effects while PCN had minimal impact. These
suggest the need for stricter PM2.5 targeted reduction policies and the
adequate resourcing of the Ghanaian health care system to improve the
early detection, management and treatment of CRDs to reduce
mortalities.

The ambient air pollutant concentrations in the urban cities were
indexed using the corresponding estimated sub-indices (Q) for each air
pollutant in the cities. The air quality index of the towns was obtained
from the worst sub-indices. The estimated air quality indices and sub-air
quality indices are provided in Table 4.

The PM2.5 sub-air quality indices (121.25 – 152.43) rated the
ambient concentrations in 56 urban cities as moderately good for human
exposure while that of Akatsi was unhealthy. The sub-indices of NO2
(2.34 – 6.09) and tropospheric O3 (34.11 – 43.74) revealed the con-
centrations in all the urban cities were good for breathing. Overall, the
air quality indices of 56 urban cities were found to be unhealthy for
sensitive groups and that of Akatsi was unhealthy for all groups of
people.

The PM2.5 sub-air quality indices per annum rated its ambient
pollution in urban cities as moderately good for breathing during the
years 2001–2003, 2005–2006, 2009–2014, and 2018–2023. However,
PM2.5 sub-air quality indices per annum rated its pollution of the
ambient air in the urban cities during the years 2000, 2004, 2007–2008,
and 2015–2017 as unhealthy for breathing. Furthermore, NO2 and O3
sub-air quality indices per year indicated their pollution in the urban
cities during the years 2000–2023 to be good for breathing. The years
2000, 2004, 2007 – 2008, and 2015 – 2017 had air quality indices
indicating the ambient air of Ghanaian urban cities was unhealthy for
breathing while that of the other years in this study’s duration was
unhealthy for sensitive groups (Table S20).

3.2. Carcinogenic, non-carcinogenic, relative and excess health risk
assessments of ambient air pollutant concentrations in urban cities

The Hazard Quotients (HQ) for one-hour exposure in the urban cities
were < 1 for all demographics, ranging from 0.49 – 0.50 for PM2.5, 0.058
– 0.059 for O3 and 0.032 – 0.033 for NO2. The hazard indices (HI) [0.58
– 0.59] for the different age groups were below the threshold of 1,
denoting no significant non-carcinogenic risks from one hour of expo-
sure to the air pollutants. HQ values ranging from 1.47 – 1.49 for PM2.5,
0.096 – 0.099 for NO2 and 0.175 – 0.177 for O3 were obtained for three
hours of exposure across all age groups. Non-carcinogenic HI of

1.73–1.75 indicated significant risks for three-hour exposure to ambient
air pollution in the cities. For six hours of exposure to PM2.5 HQ values of
2.94–2.97 were observed whereas HQ values of 0.19 – 0.20 and 0.350 –
0.353 were obtained for NO2 and O3 respectively. Non-carcinogenic HI
ranging from 1.18 – 3.51 revealed significant non-cancer risks for six-
hour exposure. Nine-hour exposure had a non-carcinogenic HQ of
4.41 – 4.46 for PM2.5, 0.288 – 0.295 NO2 and 0.52 – 0.53 for O3 expo-
sure. Non-carcinogenic HI of 5.20–5.26 was obtained for nine-hour
exposure to the air pollutants. Worst-case scenario assessments gave
HQ values of 5.88 – 5.94 for half-day and 11.76 – 11.88 for full-day adult
ambient exposure to PM2.5. The HQ values for O3 and NO2 half-day
exposure were < 1. HQ values > 1 (1.39 – 1.41) were obtained for
full-day ambient exposure to O3 while the HQ values of 0.77 – 0.79 were
obtained for NO2. Significant non-carcinogenic deleterious risks with HI
ranging from 6.94 – 7.01 for half-day and 13.96 – 14.03 for full-day
exposure were observed in this study for all the different age groups
(Table S21).

Slight differences were observed between the different age groups
regarding potential non-carcinogenic risks to ambient air pollutants. The
non-carcinogenic risks from both PM2.5 and NO2 exposure in the urban
cities could be ranked in descending order as Toddlers > Children
> Infants > Adults > Adolescents while that for O3 was Children
>Toddlers > Infants > Adults > Adolescents. This suggested that
younger age groups such as children and toddlers are more susceptible
to adverse non-cancer effects from urban air pollution in Ghana. The
individual contribution of the air pollutants to non-cancer risks in this
study could be ranked in descending order as PM2.5 > O3 > NO2. The
findings of the study suggest that traders, auto-artisans, farmers, ma-
sons, traffic wardens, street hawkers and other outdoor workers who
quotidianly spend three to nine hours in ambient environs performing
their occupational activities in urban cities in the country are more
vulnerable to non-cancer health risks. Furthermore, the exposure of
sensitive groups (such as people with pre-existing lung and cardiovas-
cular diseases) for three hours and beyond could lead to the exacerba-
tion of pre-existing health conditions.

The carcinogenic risks (CR) for human exposure to PM2.5 concen-
trations in urban cities ranged from 2.82E-06–6.76E-05 for adults,
3.95E-06–9.48E-05 for adolescents, 1.44E-05–3.44E-04 for children,
1.13E-04–2.72E-03 for infants and 3.67E-05–8.80E-04 for toddlers.
These indicated acceptable carcinogenic risks for adults, adolescents,
children and toddlers as the CR were within the tolerable carcinogenic
risk limit of 1E-04–1E-06. Contrarily, the CR for infants’ exposure to
ambient PM2.5 concentrations in urban cities indicated significant
carcinogenic risks CR for an exposure time of nine hours and beyond
(Table S22).

The estimated relative and excess risks obtained for the concentra-
tions of ambient air pollutants in the urban cities are provided in
Table S23. The relative risks for PM2.5 and O3 in the 57 urban cities in
the study were > 1 and the respective excess risks were positive. These
both suggested concentrations of PM2.5 and O3 in the ambient air of the
cities posed significant health risks to the dwellers thereof. Except for
Wa, Walewale, Tamale, Navrongo, Garu, Gushiegu, Bolgatanga, Bawku,
Accra, Kumasi and Nalerigu the relative risks for NO2 < 1 and the cor-
responding excess risks were negative. These both indicated ambient
concentrations of NO2 in most of the urban cities posed no significant
health risks. The excess risks for PM2.5, O3 and NO2 in the urban cities
varied from 0.0176 – 0.0242, 0.0059–0.0158 and (-) 0.0082 to (+)
0.0028 respectively.

Furthermore, the relative risks for PM2.5 and O3 for annual concen-
trations per annum were > 1 which indicated they posed significant
risks to the health of urban habitants in Ghana from 2000 – 2023. In
contrast, the relative risks for concentrations of NO2 per year were < 1
suggesting no significant health risks from its exposure. The excess risks
for PM2.5, O3 and NO2 per annum varied from 0.0160 to 0.0299,
0.0030–0.0222 and (-) 0.0024 to (-) 0.0046 respectively (Table S24).

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

Table 4
Classification of air quality index of ghanaian urban cities.

City QPM2.5 QO3 QNO2 AQI
Accra 142.19 34.11 5.56 Unhealthy for Sensitive Groups
Agona Swedru 145.24 34.50 3.14 Unhealthy for Sensitive Groups
Akatsi 152.43 37.07 3.49 Unhealthy
Akim Oda 143.96 37.83 3.34 Unhealthy for Sensitive Groups
Asamankese 146.10 38.07 2.48 Unhealthy for Sensitive Groups
Assin Fosu 148.82 35.34 3.08 Unhealthy for Sensitive Groups
Atebubu 143.17 42.21 3.00 Unhealthy for Sensitive Groups
Bawku 135.87 40.59 5.74 Unhealthy for Sensitive Groups
Berekum 130.95 39.68 3.95 Unhealthy for Sensitive Groups
Bimbila 141.68 43.46 4.23 Unhealthy for Sensitive Groups
Bole 131.84 41.94 3.97 Unhealthy for Sensitive Groups
Bolgatanga 132.17 40.83 5.57 Unhealthy for Sensitive Groups
Cape Coast 139.90 - 1.83 Unhealthy for Sensitive Groups
Dambai 148.35 42.59 3.69 Unhealthy for Sensitive Groups
Dormaa Ahenkro 134.73 38.18 3.42 Unhealthy for Sensitive Groups
Dunkwa Offin 140.46 36.95 2.76 Unhealthy for Sensitive Groups
Ejura 139.94 41.39 3.65 Unhealthy for Sensitive Groups
Enchi 138.25 34.48 2.36 Unhealthy for Sensitive Groups
Garu 135.58 41.32 6.09 Unhealthy for Sensitive Groups
Gushiegu 132.05 42.94 5.03 Unhealthy for Sensitive Groups
Ho 146.42 39.44 2.95 Unhealthy for Sensitive Groups
Hohoe 147.72 41.34 2.66 Unhealthy for Sensitive Groups
Jirapa 121.25 40.35 4.56 Unhealthy for Sensitive Groups
Karaga 128.53 42.79 4.90 Unhealthy for Sensitive Groups
Kete Krachi 143.13 42.56 3.28 Unhealthy for Sensitive Groups
Kintampo 135.91 41.62 2.98 Unhealthy for Sensitive Groups
Koforidua 141.62 38.31 3.54 Unhealthy for Sensitive Groups
Konongo 139.24 38.89 3.51 Unhealthy for Sensitive Groups
Kumasi 145.57 39.41 5.16 Unhealthy for Sensitive Groups
Kwame Danso 143.74 41.94 3.09 Unhealthy for Sensitive Groups
Mampong 137.03 40.54 3.15 Unhealthy for Sensitive Groups
Nalerigu 129.14 41.33 5.81 Unhealthy for Sensitive Groups
Navrongo 132.45 40.47 4.92 Unhealthy for Sensitive Groups
New Tafo Akim 139.43 39.00 3.24 Unhealthy for Sensitive Groups
Nkawkaw 140.22 39.44 3.25 Unhealthy for Sensitive Groups
Nkoranza 134.97 40.78 2.93 Unhealthy for Sensitive Groups
Nsawam 143.98 38.91 4.04 Unhealthy for Sensitive Groups
Obuasi 142.19 36.61 3.75 Unhealthy for Sensitive Groups
Paga 130.56 40.19 4.44 Unhealthy for Sensitive Groups
Salaga 143.37 42.23 3.59 Unhealthy for Sensitive Groups
Sandema 131.17 40.57 4.46 Unhealthy for Sensitive Groups
Savelugu 130.71 42.68 4.56 Unhealthy for Sensitive Groups
Sawla 131.36 42.12 3.77 Unhealthy for Sensitive Groups
Somanya 140.38 38.47 3.05 Unhealthy for Sensitive Groups
Suhum 143.49 37.99 2.82 Unhealthy for Sensitive Groups
Sunyani 133.98 39.66 4.58 Unhealthy for Sensitive Groups
Sekondi-Takoradi 137.94 - 2.34 Unhealthy for Sensitive Groups
Tamale 133.88 43.19 5.25 Unhealthy for Sensitive Groups
Tatale 134.87 43.74 4.12 Unhealthy for Sensitive Groups
Techiman 133.82 40.77 3.69 Unhealthy for Sensitive Groups
Wa 122.78 40.99 5.33 Unhealthy for Sensitive Groups
Walewale 130.20 41.69 5.10 Unhealthy for Sensitive Groups
Winneba 146.71 34.76 2.92 Unhealthy for Sensitive Groups
Wulensi 146.42 43.10 4.00 Unhealthy for Sensitive Groups
Yeji 142.62 42.45 3.66 Unhealthy for Sensitive Groups
Yendi 131.05 43.56 4.88 Unhealthy for Sensitive Groups
Zebilla 132.49 40.86 4.92 Unhealthy for Sensitive Groups

Where Q = sub-air quality index, AQI = Air Quality Index, Green Colour = Good, Orange Colour = Unhealthy for Sensitive Groups and Red Colour = Unhealthy,

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

3.3. Limitations

A notable limitation of this study is that the impact of seasonal
regional dust storms on the concentrations of pollutants for the study’s
duration was not assessed. This is due to the lack of ground and satellite
harmattan and non-harmattan pollutant measurements in the urban
cities. Furthermore, the lack of data on VOCs concentrations limited the
study from assessing the ratio of VOC to NOx to shed further light on
potential sources of ground-level ozone in the cities for the two decades.
Moreover, the concentrations of pollutants in the study are based on
satellite measurements, and the dearth of ground measurements of the
pollutants for the study’s duration prevented the study from comparing
these satellite measurements with annual ground-monitored concen-
trations. In addition, the forecasting of air pollution in the study did not
account for the impact of the COVID-19 lockdowns on air pollution
trends in the country.

3.4. Policy suggestions

Efficient ground monitoring and reporting of real-time urban air
quality along with strict enforcement of air quality standards as outlined
by the WHO and Ghana’s EPA are needed to safeguard the public’s
health. Rapid transition to renewable energy, promotion of electric ve-
hicles, improved public transportation and stricter emissions controls
and incentives are essential globally, including in Ghana.

Effective air quality management plans should be implemented in
Ghana’s urban cities. These should include reducing emissions from
transportation, industry, and waste burning. They should also promote
clean cooking technologies, enhance transportation infrastructure and
encourage ecologically conscious development. Ghana must strengthen
its regulatory framework by imposing stringent emission limits on ve-
hicles, industry and other significant polluters. Enhancing the capacity
of regulatory bodies to oversee, document and enforce adherence to air
quality guidelines is crucial.

Local governments and communities should receive training and
resources to create effective air quality management plans, alongside
international collaboration for sharing best practices and technical
support. Ghana should foster cooperation between educational in-
stitutions, public and private organizations, and businesses to develop
innovative solutions for regional air pollution focusing on health im-
pacts. National budgets, foreign donations and private sector in-
vestments should fund initiatives to improve air quality, along with
financial incentives for adopting greener practices and technologies.
Public health campaigns are needed nationwide to increase the aware-
ness of residents of urban cities on air pollution and its significant impact
on their health.

4. Conclusion

The development in urbanization, population, and industrialization
has significantly increased ambient air pollution in urban cities both
locally and globally causing some of the most serious air quality chal-
lenges. The comprehensive health risk evaluation of ambient air pollu-
tion in urban areas in Ghanaian cities from 2000 to 2023 revealed
significant findings.

The findings showed that with significant regional variations, the
yearly median PM2.5 concentrations being the most detrimental overall
for chronic respiratory diseases such as asthma, were five to six times
higher than WHO recommendations. Compared to northern cities,
Accra, Koforidua, and Kumasi consistently showed higher PM2.5 levels.
Ozone and NO2 concentrations also varied and periodically surpassed
recommended thresholds.

Most urban cities had a dominant PM2.5-O3 pollution profile, which
suggests significant health risks. Severe health risks are highlighted by
high median population attributable fractions (PAF) for air pollution
exposure especially for PM2.5. Hazard indices from PM2.5 exposure

through inhalation indicated significant non-cancer health risks for an
exposure duration of three hours and beyond per day. Carcinogenic risks
for human exposure to urban PM2.5 concentrations were mostly within
the acceptable limit of 1E-04–1E-06.

The air quality indices of 56 urban cities were found to be unhealthy
for sensitive groups and that of Akatsi was unhealthy for all groups of
people. Urban air quality indices show moderately good ambient
pollution from 2001 to 2003, 2005–2006, 2009–2014, and 2018–2023,
with unhealthy pollution from 2000 to 2017, and good sub-air quality
from 2000 to 2023. Relative risks > 1 for PM2.5 and O3 confirmed
substantial health risks for urban residents with PM2.5 pollution
contributing to a reduction in life expectancy by 4.5–6.2 years.

Regarding the severe adverse health effects of PM2.5 and O3 pollu-
tion, this study highlights the urgent need for focused actions to enhance
air quality in Ghanaian metropolitan areas. For an in-depth awareness of
air pollution dynamics in Ghana and West Africa, future studies should
prioritize bridging data gaps, including the effects of periodic dust
storms and volatile organic compounds levels.

CRediT authorship contribution statement

Marian Asantewah Nkansah: Writing – review & editing, Valida-
tion, Supervision, Project administration. Nana Kwabena Oduro
Darko: Writing – review & editing, Validation. Blessed Adjei Yeboah:
Writing – review & editing, Software, Investigation, Data curation.
Maame Serwaa Boapea: Writing – review & editing, Validation,
Investigation. Thomas Peprah Agyekum: Writing – review & editing,
Supervision, Project administration, Formal analysis, Data curation,
Conceptualization. Meshach Kojo Appiah: Writing – review & editing,
Visualization, Formal analysis. Birago Adu Ababio: Writing – review &
editing, Visualization, Formal analysis. Gerheart Winfred Ashong:
Writing – review & editing, Validation, Methodology, Formal analysis.
Lorenda Sarbeng: Writing – review & editing, Visualization, Software.
Boansi Adu Ababio: Writing – review & editing, Writing – original
draft, Visualization, Validation, Supervision, Software, Resources,
Project administration, Methodology, Investigation, Formal analysis,
Data curation, Conceptualization. Eldad Boansi: Writing – review &
editing, Validation, Conceptualization. Felix Adulley: Writing – review
& editing, Validation, Investigation. Kwabena Dabie: Writing – review
& editing, Validation, Methodology. Edward Ebow Kwaansa-Ansah:
Writing – review & editing, Validation, Supervision, Project adminis-
tration. Michael Kweku Commeh: Writing – review & editing, Super-
vision, Project administration. Jonathan Nartey Hogarh: Writing –
review & editing, Supervision, Project administration.

Declaration of Competing Interest

The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.

Acknowledgements

We thank the Milken Institute of Public Health for making their data
available. We also thank Susan Anenberg for her assistance and valuable
feedback during this work.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the
online version at doi:10.1016/j.ecoenv.2024.117591.

Data availability

Data will be made available on request.

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

https://doi.org/10.1016/j.ecoenv.2024.117591

References

Ababio, B.A., 2023. Evaluation of indoor air quality of senior high school kitchens in the
Kumasi metropolis (Masters Dissertation).

Ababio, B.A., Nkansah, M.A., Hogarh, J.N., Agyekum, T.P., Commeh, M.K., 2023. In-
kitchen particulate matter emissions in high schools in the Kumasi metropolis,
Ghana: levels and the health risk assessment. J. Hazard. Mater. Adv., 100358

Ababio, B.A., Nkansah, M.A., Hogarh, J.N., Agyekum, T.P., Commeh, M.K., 2024.
Gaseous air quality and health risk assessment of high school kitchens in the Kumasi
Metropolis. Environ. Adv. 17, 100576.

Abdul Raheem, M., Jimoh, G., Abdulrahim, H., 2022. Assessment of kitchen air
pollution: Health implications for the residents of ilorin south, nigeria. J. Environ.
Public Health 2022 (1), 7689141.

Agodzo, S.K., Bessah, E., Nyatuame, M., 2023. A review of the water resources of Ghana
in a changing climate and anthropogenic stresses. Front. Water 4, 973825.

Alli, A.S., Clark, S.N., Wang, J., Bennett, J., Hughes, A.F., Ezzati, M., Brauer, M.,
Nimo, J., Bedford-Moses, J., Baah, S., Cavanaugh, A., 2023. High-resolution patterns
and inequalities in ambient fine particle mass (PM2. 5) and black carbon (BC) in the
Greater Accra Metropolis, Ghana. Sci. Total Environ. 875, 162582.

Amegah, A.K., Dakuu, G., Mudu, P., Jaakkola, J.J., 2022. Particulate matter pollution at
traffic hotspots of Accra, Ghana: levels, exposure experiences of street traders, and
associated respiratory and cardiovascular symptoms. J. Expo. Sci. Environ.
Epidemiol. 32 (2), 333–342.

Anenberg, S.C., Mohegh, A., Goldberg, D.L., Kerr, G.H., Brauer, M., Burkart, K.,
Hystad, P., Larkin, A., Wozniak, S., Lamsal, L., 2022b. Long-term trends in urban
NO2 concentrations and associated paediatric asthma incidence: estimates from
global datasets. Lancet Planet. Health 6 (1), e49–e58.

Arku, R.E., Dionisio, K.L., Hughes, A.F., Vallarino, J., Spengler, J.D., Castro, M.C., Agyei-
Mensah, S., Ezzati, M., 2015. Personal particulate matter exposures and locations of
students in four neighborhoods in Accra, Ghana. J. Expo. Sci. Environ. Epidemiol. 25
(6), 557–566.

Arku, R.E., Vallarino, J., Dionisio, K.L., Willis, R., Choi, H., Wilson, J.G., Hemphill, C.,
Agyei-Mensah, S., Spengler, J.D., Ezzati, M., 2008. Characterizing air pollution in
two low-income neighborhoods in Accra, Ghana. Sci. Total Environ. 402 (2-3),
217–231.

Armah, F.A., Yawson, D.O., Pappoe, A.A., 2010. A systems dynamics approach to explore
traffic congestion and air pollution link in the city of Accra, Ghana. Sustainability 2
(1), 252–265.

Ashong, G.W., Ababio, B.A., Kwaansa-Ansah, E.E., Gyabeng, E., Nti, S.O., 2024. Human
and ecotoxicological risk assessment of heavy metals in polymer post treatment
sludge from Barekese Drinking Water Treatment Plant, Kumasi. Toxicol. Rep. 12,
404–413.

Azhari, A., Halim, N.D.A., Mohtar, A.A.A., Aiyub, K., Latif, M.T., Ketzel, M., 2021.
Evaluation and prediction of PM10 and PM2. 5 from road source emissions in Kuala
Lumpur City Centre. Sustainability 13 (10), 5402.

Bahino, J., Giordano, M., Beekmann, M., Yoboue, V., Ochou, A., Galy-Lacaux, C.,
Liousse, C., Hughes, A., Nimo, J., Lemmouchi, F., Cuesta, J., 2024. Temporal
variability and regional influences of PM 2.5 in the West African cities of Abidjan
(Côte d′Ivoire) and Accra (Ghana). Environ. Sci.: Atmos. 4 (4), 468–487.

Cakaj, A., Lisiak-Zielińska, M., Khaniabadi, Y.O., Sicard, P., 2023. Premature deaths
related to urban air pollution in Poland. Atmos. Environ. 301, 119723.

Chen, Y., Ebenstein, A., Greenstone, M., Li, H., 2013. Evidence on the impact of sustained
exposure to air pollution on life expectancy from China’s Huai River policy. Proc.
Natl. Acad. Sci. 110 (32), 12936–12941.

Dionisio, K.L., Arku, R.E., Hughes, A.F., Vallarino, J., Carmichael, H., Spengler, J.D.,
Agyei-Mensah, S. and Ezzati, M., 2010. Air pollution in Accra neighborhoods:
spatial, socioeconomic, and temporal patterns.

Ebenstein, A., Fan, M., Greenstone, M., He, G., Zhou, M., 2017. New evidence on the
impact of sustained exposure to air pollution on life expectancy from China’s Huai
River Policy. Proc. Natl. Acad. Sci. 114 (39), 10384–10389.

Fan, H., Zhao, C., Yang, Y., 2020. A comprehensive analysis of the spatio-temporal
variation of urban air pollution in China during 2014–2018. Atmos. Environ. 220,
117066.

Ghana, G.S.S., 2021. Population and Housing Census: Population of Regions and
Districts. Ghana Statistical Service.

Greenstone, M. and Fan, C.Q., 2018. Introducing the air quality life index. AQLI Annual
Report.

He, C., Liu, J., Zhou, Y., Zhou, J., Zhang, L., Wang, Y., Liu, L., Peng, S., 2024. Synergistic
PM2. 5 and O3 control to address the emerging global PM2. 5-O3 compound
pollution challenges. Eco-Environ. Health. https://doi.org/10.1016/j.
eehl.2024.04.004.

He, C., Wu, Q., Li, B., Liu, J., Gong, X., Zhang, L., 2023. Surface ozone pollution in China:
trends, exposure risks, and drivers. Front. Public Health 11, 1131753.

Hogarh, J.N., Agyekum, T.P., Bempah, C.K., Owusu-Ansah, E.D.J., Avicor, S.W.,
Awandare, G.A., Fobil, J.N., Obiri-Danso, K., 2018. Environmental health risks and
benefits of the use of mosquito coils as malaria prevention and control strategy.
Malar. J. 17 (1), 1–12. https://doi.org/10.1186/s12936-018-2412-4.

Huang, J., Pan, X., Guo, X., Li, G., 2018. Health impact of China’s air pollution
prevention and control action plan: an analysis of national air quality monitoring
and mortality data. Lancet Planet. Health 2 (7), e313–e323.

Iddrisu, S., Siiba, A., Alhassan, J., Abass, K., 2023. Land-use and land cover change
dynamics in urban Ghana: implications for peri-urban livelihoods. Int. J. Urban
Sustain. Dev. 15 (1), 80–96.

Johnson, M., Mazur, L., Fisher, M., Fraser, W.D., Sun, L., Hystad, P., Gandhi, C.K., 2024.
Prenatal exposure to air pollution and respiratory distress in term newborns: results

from the MIREC prospective pregnancy cohort. Environ. Health Perspect. 132 (1),
017007.

Jurado, X., Reiminger, N., Vazquez, J., Wemmert, C., Dufresne, M., Blond, N., Wertel, J.,
2020. Assessment of mean annual NO2 concentration based on a partial dataset.
Atmos. Environ. 221, 117087.

Kanhai, G., Fobil, J.N., Nartey, B.A., Spadaro, J.V., Mudu, P., 2021. Urban Municipal
Solid Waste management: modeling air pollution scenarios and health impacts in the
case of Accra, Ghana. Waste Manag. 123, 15–22.

Kuerban, M., Waili, Y., Fan, F., Liu, Y., Qin, W., Dore, A.J., Peng, J., Xu, W., Zhang, F.,
2020. Spatio-temporal patterns of air pollution in China from 2015 to 2018 and
implications for health risks. Environ. Pollut. 258, 113659.

Kumar, P., Singh, A.B., Arora, T., Singh, S., Singh, R., 2023. Critical review on emerging
health effects associated with the indoor air quality and its sustainable management.
Sci. Total Environ. 872, 162163.

Lee, H.J., Kuwayama, T., FitzGibbon, M., 2023. Trends of ambient O3 levels associated
with O3 precursor gases and meteorology in California: synergies from ground and
satellite observations. Remote Sens. Environ. 284, 113358.

Li, C., van Donkelaar, A., Hammer, M.S., McDuffie, E.E., Burnett, R.T., Spadaro, J.V.,
Chatterjee, D., Cohen, A.J., Apte, J.S., Southerland, V.A., Anenberg, S.C., 2023.
Reversal of trends in global fine particulate matter air pollution. Nat. Commun. 14
(1), 5349.

Li, J., Garshick, E., Hart, J.E., Li, L., Shi, L., Al-Hemoud, A., Huang, S., Koutrakis, P.,
2021. Estimation of ambient PM2. 5 in Iraq and Kuwait from 2001 to 2018 using
machine learning and remote sensing. Environ. Int. 151, 106445.

Liu, X., Yi, G., Zhou, X., Zhang, T., Bie, X., Li, J., Tan, H., 2023. Spatio-temporal
variations of PM2. 5 and O3 in China during 2013–2021: impact factor analysis.
Environ. Pollut. 334, 122189.

Ma, Y.H., Chen, H.S., Liu, C., Feng, Q.S., Feng, L., Zhang, Y.R., Hu, H., Dong, Q., Tan, L.,
Kan, H.D., Zhang, C., 2023. Association of long-term exposure to ambient air
pollution with cognitive decline and Alzheimer’s disease–related amyloidosis. Biol.
Psychiatry 93 (9), 780–789.

Malashock, D.A., Delang, M.N., Becker, J.S., Serre, M.L., West, J.J., Chang, K.L.,
Cooper, O.R., Anenberg, S.C., 2022. Global trends in ozone concentration and
attributable mortality for urban, peri-urban, and rural areas between 2000 and 2019:
a modelling study. Lancet Planet. Health 6 (12), e958–e967.

Mandal, S., Madhipatla, K.K., Guttikunda, S., Kloog, I., Prabhakaran, D., Schwartz, J.D.,
Team, G.H.I., 2020. Ensemble averaging based assessment of spatiotemporal
variations in ambient PM2. 5 concentrations over Delhi, India, during 2010–2016.
Atmos. Environ. 224, 117309.

Mbazima, S.J., 2022. Health risk assessment of particulate matter 2.5 in an academic
metallurgy workshop. Indoor Air 32 (9), e13111.

McHugh, K., Cummins, T., Aherne, J., 2023. Distribution and long-term trends of
tropospheric ozone concentrations in Ireland. Atmosphere 14 (3), 569.

Min, J., Kang, D.H., Kang, C., Bell, M.L., Kim, H., Yang, J., Gasparrini, A., Lavigne, E.,
Hashizume, M., Kim, Y., Ng, C.F.S., 2024. Fluctuating risk of acute kidney injury-
related mortality for four weeks after exposure to air pollution: a multi-country time-
series study in 6 countries. Environ. Int. 183, 108367.

Momtazmanesh, S., Moghaddam, S.S., Ghamari, S.H., Rad, E.M., Rezaei, N., Shobeiri, P.,
Aali, A., Abbasi-Kangevari, M., Abbasi-Kangevari, Z., Abdelmasseh, M., Abdoun, M.,
2023. Global burden of chronic respiratory diseases and risk factors, 1990–2019: an
update from the Global Burden of Disease Study 2019. EClinicalMedicine 59.

Moro, A., Aborigo, R., Nonterah, E., Seyram, K., Nartey, K. and Soyiri, I., 2022,
September. Variation in PM2. 5, PM10 and atmospheric conditions in the three (3)
ecological zones in Ghana. In ISEE Conference Abstracts (Vol. 2022, No. 1).

Nyadanu, S.D., Tessema, G.A., Mullins, B., Kumi-Boateng, B., Ofosu, A.A., Pereira, G.,
2022. Ambient particulate matter air pollution and stillbirth in Ghana: a difference-
in-differences approach. Atmos. Pollut. Res. 13 (7), 101471.

Nyarku, M., Buonanno, G., Ofosu, F., Jayaratne, R., Mazaheri, M., Morawska, L., 2021.
Ultrafine particles in key microenvironments in rural and urban areas of Ghana.
Atmos. Pollut. Res. 12 (11), 101212.

Odonkor, S.T., Mahami, T., 2020. Knowledge, attitudes, and perceptions of air pollution
in Accra, Ghana: a critical survey. J. Environ. Public Health.

Ofosu, F.G., Hopke, P.K., Aboh, I.J., Bamford, S.A., 2012. Characterization of fine
particulate sources at Ashaiman in Greater Accra, Ghana. Atmos. Pollut. Res. 3 (3),
301–310.

Okedere, O.B., Elehinafe, F.B., Oyelami, S., Ayeni, A.O., 2021. Drivers of anthropogenic
air emissions in Nigeria-a review. Heliyon 7 (3).

Park, E.H., Heo, J., Kim, H., Yi, S.M., 2020. Long term trends of chemical constituents
and source contributions of PM2. 5 in Seoul. Chemosphere 251, 126371.

Rahman, M.S., Bhuiyan, S.S., Ahmed, Z., Saha, N., Begum, B.A., 2021. Characterization
and source apportionment of elemental species in PM2. 5 with especial emphasis on
seasonal variation in the capital city “Dhaka”, Bangladesh. Urban Clim. 36, 100804.

Rana, S., 2022. Determination of Air Quality Life Index (AQLI) In Medinipur City of West
Bengal (India) during 2019 to 2020: a contextual study. Curr. World Environ. 12 (1).

Rooney, M.S., Arku, R.E., Dionisio, K.L., Paciorek, C., Friedman, A.B., Carmichael, H.,
Zhou, Z., Hughes, A.F., Vallarino, J., Agyei-Mensah, S., Spengler, J.D., 2012. Spatial
and temporal patterns of particulate matter sources and pollution in four
communities in Accra, Ghana. Sci. Total Environ. 435, 107–114.

Sanyal, S., Rochereau, T., Maesano, C.N., Com-Ruelle, L., Annesi-Maesano, I., 2018.
Long-term effect of outdoor air pollution on mortality and morbidity: a 12-year
follow-up study for metropolitan France. Int. J. Environ. Res. Public Health 15 (11),
2487.

Sharma, K., Garg, A., Joshi, V., Kumar, A., 2023. Assessment of health risks for criteria
air pollutants present in 11 non-attainment cities of Uttar Pradesh, India. Hum. Ecol.
Risk Assess.: Int. J. 29 (1), 103–122.

B.A. Ababio et al. Ecotoxicology and Environmental Safety 289 (2025) 117591

http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref1
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref1
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref1
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref2
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref2
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref2
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref3
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref3
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref3
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref4
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref4
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref5
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref5
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref5
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref5
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref6
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref6
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref6
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref6
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref7
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref7
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref7
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref7
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref8
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref8
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref8
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref8
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref9
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref9
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref9
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref9
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref10
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref10
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref10
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref11
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref11
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref11
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref11
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref12
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref12
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref12
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref13
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref13
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref13
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref13
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref14
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref14
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref15
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref15
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref15
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref16
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref16
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref16
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref17
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref17
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref17
https://doi.org/10.1016/j.eehl.2024.04.004
https://doi.org/10.1016/j.eehl.2024.04.004
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref19
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref19
https://doi.org/10.1186/s12936-018-2412-4
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref21
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref21
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref21
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref22
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref22
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref22
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref23
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref23
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref23
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref23
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref24
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref24
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref24
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref25
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref25
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref25
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref26
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref26
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref26
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref27
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref27
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref27
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref28
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref28
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref28
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref29
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref29
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref29
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref29
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref30
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref30
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref30
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref31
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref31
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref31
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref32
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref32
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref32
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref32
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref33
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref33
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref33
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref33
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref34
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref34
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref34
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref34
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref35
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref35
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref36
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref36
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref37
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref37
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref37
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref37
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref38
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref38
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref38
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref38
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref39
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref39
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref39
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref40
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref40
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref40
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref41
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref41
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref42
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref42
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref42
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref43
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref43
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref44
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref44
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref45
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref45
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref45
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref46
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref46
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref47
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref47
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref47
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref47
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref48
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref48
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref48
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref48
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref49
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref49
http://refhub.elsevier.com/S0147-6513(24)01667-1/sbref49

Shen, L.T., Ge, M.W., Hu, F.H., Jia, Y.J., Tang, W., Zhang, W.Q., Zhao, D.Y., Shen, W.Q.,
Chen, H.L., 2023. The connection between six common air pollution particles and
adult brain tumors: a meta-analysis of 26,217,930 individuals. Environ. Sci. Pollut.
Res. 30 (50), 108525–108537.

Shen, F., Zhang, L., Jiang, L., Tang, M., Gai, X., Chen, M., Ge, X., 2020. Temporal
variations of six ambient criteria air pollutants from 2015 to 2018, their spatial
distributions, health risks and relationships with socioeconomic factors during 2018
in China. Environ. Int. 137, 105556.

Sicard, P., Agathokleous, E., Anenberg, S.C., De Marco, A., Paoletti, E., Calatayud, V.,
2023. Trends in urban air pollution over the last two decades: a global perspective.
Sci. Total Environ. 858, 160064.

Sicard, P., Agathokleous, E., De Marco, A., Paoletti, E., Calatayud, V., 2021. Urban
population exposure to air pollution in Europe over the last decades. Environ. Sci.
Eur. 33, 28.

Sicard, P., Paoletti, E., Agathokleous, E., Araminienė, V., Proietti, C., Coulibaly, F., De
Marco, A., 2020. Ozone weekend effect in cities: deep insights for urban air pollution
control. Environ. Res. 191, 110193.

Southerland, V.A., Brauer, M., Mohegh, A., Hammer, M.S., Van Donkelaar, A., Martin, R.
V., Apte, J.S., Anenberg, S.C., 2022. Global urban temporal trends in fine particulate
matter (PM2⋅ 5) and attributable health burdens: estimates from global datasets.
Lancet Planet. Health 6 (2), e139–e146.

Tariq, S., Mariam, A., Mehmood, U., 2023. Assessment of variability in PM2. 5 and its
impact on human health in a West African country. Chemosphere 344, 140357.

United States Environmental Protection Agency, 2011. Exposure factors handbook: 2011
Edition. Office of Research and Development, Washington, DC.

VoPham, T., Jones, R.R., 2023. State of the science on outdoor air pollution exposure and
liver cancer risk. Environ. Adv., 100354

Wang, J., Alli, A.S., Clark, S., Hughes, A., Ezzati, M., Beddows, A., Vallarino, J., Nimo, J.,
Bedford-Moses, J., Baah, S., Owusu, G., 2022. Nitrogen oxides (NO and NO2)
pollution in the Accra metropolis: spatiotemporal patterns and the role of
meteorology. Sci. Total Environ. 803, 149931.

Wang, J., Alli, A.S., Clark, S.N., Ezzati, M., Brauer, M., Hughes, A.F., Nimo, J., Bedford-
Moses, J., Baah, S., Nathvani, R. and Agyei-Mensah, S., 2024. Inequalities in urban
air pollution in sub-Saharan Africa: An empirical modelling of ambient NO and NO2
concentrations in Accra, Ghana. Environmental Research Letters.

Wang, R., Liu, J., Qin, Y., Chen, Z., Li, J., Guo, P., Shan, L., Li, Y., Hao, Y., Jiao, M., Qi, X.,
2023. Global attributed burden of death for air pollution: demographic
decomposition and birth cohort effect. Sci. Total Environ. 860, 160444.

World Bank Group. 2020. Ghana Country Environmental Analysis. Washington, D.C. 〈htt
p://documents.worldbank.org/curated/en/419871588578973802/Ghana-Countr
y-Environmental-Analysis〉 p. xix.

World Health Organization, 2021. . WHO global air quality guidelines: particulate matter
(PM2. 5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide.
World Health Organization.

Xu, X., Zhang, T., Su, Y., 2019. Temporal variations and trend of ground-level ozone
based on long-term m