Neonatal Mortality Rate in the Context of Air Pollution: A Comparative Investigation

Abstract

This study investigates the relationship between carbon dioxide (CO₂) emissions and inhalable particulate matter levels with a diameter of at most 2.5 micrometers (PM2.5) on neonatal mortality rates across 88 countries, building upon an extensive amount of literature highlighting the harmful effects of air pollution on health. We categorize the countries based on their levels of CO₂ emissions and PM2.5 pollution as high versus low emitters, showing a statistically significant disparity in the average neonatal mortality rates between countries with high and low levels of CO₂ and PM2.5. Further exploring the underlying factors influencing the neonatal mortality rate within each group with the help of regression analysis, we identified several significant socioeconomic, environmental, and health-related factors affecting the neonatal mortality rate. Our findings highlight urgent public health concerns toward the achievement of the Sustainable Development Goals, particularly SDG 3, at a country level that address health-related issues in varying contexts of environmental pollution.

1. Introduction

Researchers have found that exposure to air pollution is linked to a variety of detrimental health consequences, such as alterations in cardiovascular and lung function, ER visits, hospital stays, and early death [1,2]. A study by [3] examined the impacts of environmental pollution on human health by looking at a panel dataset that covered 23 years. Their results confirm that public health in Asia’s developing nations is significantly impacted by carbon emissions. A review study by [4] also highlighted that climate variables, particularly temperature, and air pollutants had a significant influence on human health. Studies also highlighted that in many cities, particulate matter smaller than 2.5 μm has emerged as the main cause of air pollution, and this term is frequently used to describe air pollution [5]. PM2.5’s small size allows it to enter the respiratory system and then pass through the bronchi. Many epidemiological studies have found that hospitalizations for respiratory and cardiovascular conditions increase with both short-term and long-term exposure to PM2.5, suggesting that PM2.5 can have both immediate and long-term health impacts [6]. A similar study by [7] assessed the link between PM2.5 exposure and cause-specific death in this investigation. They reported the relationship between PM2.5 exposure and mortality from lung cancer, pneumonia, COPD, type 2 diabetes, cardiovascular disease, and cerebrovascular illness. Similar research adds dementia, hypertension, and chronic renal disease to the list of recognized causes of death linked to PM2.5 exposure [8]. Another nationwide investigation discovered strong correlations between eight PM2.5 components that naturally cause mortality. They added that mineral dust and elemental carbon were strongly linked to increased mortality from lung and respiratory cancers [9]. A study by [10] discovered that a 1 μg/m³ rise in PM2.5 resulted in a 3.85 × 10⁻⁴ increase in the likelihood of dying that year. A similar study by [11] discovered that the number of children dying before the age of five increases by around 14.5% for every unit annual increase in PM2.5, indicating the seriousness of PM2.5’s health impacts.

In addition, industry revolution and higher energy consumption tend to increase CO₂ emissions, particularly when it is powered by fossil fuels. Researchers have discovered that vulnerable groups, such as children, the elderly, and pregnant women, are most affected by the harmful health impacts of CO₂ emissions [12]. Researchers claim that continuous exposure to atmospheric CO₂ could be an unnoticed stressor for human beings and the future environment [13,14,15]. A recent panel study by [16] demonstrated that CO₂ emissions cause significant short- and long-term increases in baby and under-5 death rates. Additionally, researchers contend that compared to infants born in 1960, children born in 2020 are predicted to encounter two to seven times as many dangerous weather occurrences [17]. Additional reasons emphasize that children in areas with high levels of air pollution exhibit more changes in the tissue structure of their airway mucosa [18]. Comparing the nasal tissues of children living in Mexico City, a severely polluted metropolitan location, to children living in Veracruz City, a relatively cleaner place, showed notable ultrastructural differences [19]. Furthermore, research has shown that the negative effects of air pollution become apparent as soon as a child is conceived. When pregnant women are exposed to high levels of air pollution, the contaminants enter their bloodstream and travel through the placenta and umbilical cord blood to the fetus [20].

The Agenda for the Sustainable Development Goals (SDGs) involves 17 goals [21]. Countries all around the world will be required to include the SDGs in their own national development plans if this agenda is put into action. The SDGs’ interconnectedness implies that relationships with other goals frequently have an impact on how well one target is progressing [22]. A key element of the wide range of subjects addressed by the SDGs is the analysis of relationships between SDG indicators and targets. Policymakers must carefully evaluate various interactions to differentiate between possible supporters and negotiation partners in order to properly achieve the SDGs. This will enable the quick and efficient development of coherent plans [23].

To deepen our understanding of the relationship between air pollutants and neonatal mortality rate, we categorized the 88 countries into two groups based on their CO₂ and PM2.5 emission levels: high and low emitters. Despite the growing body of literature on the relationship between air pollution and human health, limited research has specifically examined the relationship between neonatal mortality rates and air pollutants across two groups of countries with high and low levels of emissions. Studying air pollution and neonatal mortality is vital because, as shown in multiple studies by other researchers, prenatal exposure to pollutants increases the risks of low birth weight, preterm birth, and neonatal death globally. The integration of several SDGs, e.g., SDG 3 (Good Health and Well-Being) and SDG 11 (Sustainable Cities and Communities), as well as SDG 13 (Climate Action) and SDG 10 (Reduced Inequalities), is consistent with this study. Understanding the intricate relationships between mortality rate, health, and air quality is essential to creating and implementing successful policies and achieving the SDGs. SDG 3, particularly SDG 3.2, targets ending the avoidable neonatal mortality rate by 2030. Ref. [24], using data from 1981 to 2019 across South Asia and Africa, showed that CO₂ emissions are the primary barrier to achieving SDG 3.2 by 2030. Based on the availability of data from publicly accessible websites like the World Bank and the WHO, we selected data for our studies, based on 88 countries in 2022. We limited our investigation to 88 nations since we only looked at those with comprehensive data to guarantee the validity of our conclusions. Additionally, our study aims to determine the key variables influencing neonatal mortality rates in two sets of nations with high and low levels of emissions in 2022.

2. Literature Review

2.1. Air Pollution and Health

It is widely accepted that air pollution is a serious worldwide issue that endangers people’s health and lowers their quality of life. Researchers have found that a cleaner environment positively influences human development by advancing their health [25]. The most dominant air pollutants are CO₂ and PM2.5, which are frequently employed as indicators of IAQ [26]. Studies have stated that air pollution from industries like manufacturing and transportation is harmful, and that policies and methods aimed at creating low-carbon environments are necessary to reduce health concerns [27]. As discussed by [28], methane, climate, ecosystems, and human health are all positively correlated. A new perspective highlighted by [29] demonstrates that low-income people in Delhi would benefit more from improved air quality in terms of their health.

2.2. Air Pollution and Mortality Rates

In addition, scholars identified a link between CO₂ emissions and mortality rate [14,30,31]. A study by [32] assessed the prevalence of neonatal diarrheal illnesses worldwide between 1990 and 2019 that are linked to PM2.5 air pollution. They argue that about 10,386 infant deaths and a considerable number of disability-adjusted life years (DALYs) can be attributed to PM2.5. A similar study by [33] reported that since 1990, the prevalence of PM2.5-related non-communicable diseases has grown, especially in South Asia, Sub-Saharan Africa, and low–middle-sociodemographic-index regions. The average annual percentage change for the worldwide age-standardized mortality rate in 2019 was 0.91, with 2.09 deaths per 100,000 people. As highlighted by [34], long-term exposure to ambient PM2.5 poses a serious risk to global health and that child mortality is still high in some nations; their findings emphasized the need for more research in this area. According to some other study findings, perinatal death rates and PM2.5 are substantially positively correlated. Perinatal mortality rates rise by 1.76 and 2.31 percent for every 1% increase in the log-transformed average and maximum concentrations of PM2.5, respectively. They also added that perinatal death rates rise by 2.49 and 2.19 percent, respectively, for every 1% increase in log-transformed average and maximum PM2.5 concentrations [32]. Furthermore, other study findings show a 10% rise in the PM2.5 level results in a mortality rate boost of 0.04% to 0.06% among adults [31]. Studies of mortality caused by fine particulate matter in the world’s 250 most populous cities indicate that city-level PM2.5-attributable mortality rates varied from 13 to 125 deaths per 100,000 people [35].

An investigation by [36] showed that the long-term infant death rate is significantly influenced by air pollution, healthcare costs, and economic factors. In a similar vein, a recent study that examined four distinct urban design patterns in European cities found that low-density cities have significantly less air pollution and lower mortality rates than other urban design types [37]. According to a Commonwealth of Independent States study, CO₂ emissions are responsible for the largest variations in death rates because they cause diabetes, cardiovascular disease, and chronic respiratory conditions [38]. Higher levels of PM10 and CO₂ have a significant impact on increased death rates among newborns, very young children, and adults, according to a similar analysis conducted using a panel dataset that includes 35 African countries. As highlighted by [39], the detrimental effects of CO₂ emissions on newborns and recommends that CO₂ levels be regularly checked after delivery, during transportation, and for infants receiving invasive or non-invasive respiratory support in the NICU. According to a different African study that looked at the relationship between CO₂ emissions and the under-five mortality rate, if all other parameters remained the same, a 1% increase in the use of fossil fuels would lead to a 0.418% increase in the under-five mortality rate [40]. According to a 2005–2015 panel analysis of 42 Asian and Pacific nations, CO₂ emissions are linked to a higher newborn death rate. Furthermore, the results of this study show that CO₂ emissions exhibit variable mediation effects for the various income categories. For instance, the CO₂ emissions mediation effect in lower- and upper-middle-income nations is comparable to that of the entire sample [41]. In a similar vein, an Azerbaijani study emphasizes the serious harm that environmental pollution causes to newborns’ health. They found that newborn death rates increase by an average of 1.69% for every 1% increase in CO₂ emissions [42]. A recent study by [43] reviews the links between air pollution and health, focusing on SDG 3. It shows links between air pollution and maternal and child health and emphasizes the importance of reducing air pollution to achieve SDG 3. As lined out by [44], in order to successfully reduce the number of deaths caused by PM2.5 pollution, SDG 3.9 calls for additional efforts to be made in the areas of healthcare facilities and air pollution reduction.

2.3. Socioeconomic Factors and Mortality Rate

In addition, a Pakistani study found that economic factors have a greater impact on child mortality rates than CO₂ emissions. They revealed that child mortality is higher in lower-income homes, primarily as a result of poor living circumstances and limited access to healthcare [45]. A study by [41] also discovered a strong inverse relationship between under-5 death rates and income. They contend that the beneficial impact of emissions alone on mortality is in fact eliminated by the relationship between emissions and per capita income. According to a recent study, household air pollution is more likely to be the cause of high infant mortality rates in low-income nations [8]. Of all the research studies that looked at the relationship between air pollution, health, and mortality rates, a number of them found that socioeconomic status, education level, the level of the gender inequality index (GII), economic growth, and rurality had a major impact on the risk of children’s health and neonatal deaths, and they highlighted that socioeconomic factors and economic growth may indirectly affect health, mainly by raising CO₂ levels [46,47,48]. Additionally, a number of earlier research demonstrated a strong and adverse correlation between public healthcare spending and the rates of infant and neonatal mortality [49]. According to a related study, there was a −0.635 variance in the neonatal mortality rate for every 1% change in the overall amount of public spending in the health sector and other sectors [50].

The present research aims to investigate the factors affecting the neonatal mortality rate in regions with high and low levels of CO₂ and PM2.5 emissions in 2022. By examining data from diverse geographical regions, this study seeks to discover the possible link between levels of air pollution and the extent of the neonatal mortality rate. This study also looks at a number of environmental and socioeconomic factors that might influence these results. The SDGs, such as Goal 3—“ensuring healthy lives and promoting well-being for all at all ages”—and Goal 11—“making cities and human settlements inclusive, safe, resilient, and sustainable”—are in line with our research. In order to safeguard vulnerable groups, especially infants and future generations, we seek to educate legislators and medical professionals on the important variables influencing the neonatal mortality rate.

3. Materials and Methods

The present study is based on the most recent publicly available data provided by the World Bank [51] and the WHO [52] regarding CO₂ emissions, PM2.5 air pollution, the neonatal mortality rate, and environmental, health, and economic factors related to the SDGs. Complete data are available for 88 countries, and the factors studied for every country are listed and defined in

Table 1. Definition of the environmental, health, and economic factors related to the SDGs (independent variables).

Variable	Indicator Name	Definition
X1	Gender inequality index (GII)	It assesses disparities between women and men by examining factors such as reproductive health outcomes (maternal mortality rates, adolescent birth rates), empowerment indicators (access to at least some secondary education, proportion of parliamentary seats held by women), and women’s participation in the labor market. The scores range from 0 to 1, with lower values signifying less disparity between women and men.
X2	Women, business and the law index score (scale 1–100)	It assesses the impact of laws and regulations on women’s economic opportunities by averaging the score of each of eight indices, mobility, workplace, pay, marriage, parenthood, entrepreneurship, assets, and pension, with a maximum score of 100.
X3	Ratio of female to male labor force participation rate (%) (modeled ILO estimate)	The percentage of the population aged 15 and above that is engaged in economic activity, encompassing all individuals who provide labor for the production of goods and services over a designated period. It is determined by dividing the female labor force participation rate by the male labor force participation rate and then multiplying by 100.
X4	Current health expenditure per capita, PPP (current international USD)	Average spending on healthcare per person, measured in international dollars adjusted for local purchasing power.
X5	Hospital beds (per 1000 people)	Hospital beds for both acute and chronic care available in public, private, general, and specialized hospitals, and in rehabilitation centers.
X6	People using at least basic sanitation services (% of population)	Improved sanitation facilities (flush and pour-flush connections to piped sewer systems, septic tanks, or pit latrines; ventilated pit latrines, composting toilets, or pit latrines with concrete slabs) not shared with other households, considering individuals utilizing basic sanitation facilities, in addition to those who have access to safely managed sanitation services.
X7	Forest area (% of land area)	Land covered by natural or cultivated tree stands at least 5 m in height is considered forested, regardless of its productivity, excluding areas with trees used for agricultural purposes (such as fruit plantations and agroforestry systems) and trees in urban green spaces.
X8	Crop production index (2014–2016 = 100)	It displays agricultural output for each year compared to a base period spanning 2014–2016, excluding crops that are grown for animal feed.
X9	Rural population (% of total population)	The percentage of people living in rural areas, calculated using the difference between the total and the urban population.

. Carbon dioxide is one of the major greenhouse gases contributing to global warming. Annual CO₂ emissions from agriculture, energy, waste, and industry per capita are measured in tonnes of CO₂ equivalent (t CO₂e). This measure excludes GHG emissions resulting from land use changes, land use, and forestry activities. PM2.5 is one of the components of primary fine particulate matter and of the main air pollutants. The mean annual population-weighted exposure to ambient PM2.5 pollution is the average level of exposure of the population to concentrations of suspended particles measuring less than 2.5 microns in aerodynamic diameter measured in micrograms per cubic meter.

First, we explore the distribution of variables and analyze the descriptive statistics for all countries. As we study the neonatal mortality rate in the context of air pollution, we focus on analyzing CO₂ emissions and PM2.5 pollution. In order to study the associations between environmental, health, and economic factors listed in Table 1 and the neonatal mortality rate, we identify countries with low versus high CO₂ emissions and countries with low versus high PM2.5 pollution. The thresholds are calculated as the median value of CO₂ emissions and PM2.5 pollution. We test whether there is a statistically significant difference in the average CO₂ emissions and in the average PM2.5 pollution in the two groups of countries. We also test whether the average neonatal mortality rate is significantly different in countries with low CO₂ emissions versus in countries with high CO₂ emissions, and in countries with low vs. high PM2.5 pollution. Finally, we build four multiple regression models for these four groups of countries using the variables from Table 1 and evaluate each model. We compare the regression models in terms of factors that exhibit a significant association with the neonatal mortality rate and select the best model that is statistically significant, with the highest proportion of variance in the neonatal mortality rate that can be explained by the independent variables.

4. Results

Descriptive statistics are summarized in

Table 2. Descriptive statistics of the variables, including the minimum, maximum, average, median, standard deviation, kurtosis, and skewness.

	Y	CO2	PM2.5	X1	X2	X3	X4	X5	X6	X7	X8	X9
Minimum	0.79	0.11	4.90	0.01	46.88	21.74	56.67	1.98	14.75	0.52	53.79	1.85
Maximum	28.64	15.11	63.74	0.65	100.00	98.35	8998.39	127.20	100.00	73.73	189.93	82.28
Mean	9.11	3.88	20.54	0.28	84.79	76.32	2341.42	32.11	82.87	33.63	108.29	36.13
Median	6.79	3.18	17.10	0.28	86.56	81.05	1208.45	25.31	94.28	33.10	106.92	34.01
Standard Deviation	7.98	3.26	12.95	0.18	12.15	15.01	2469.43	25.02	23.26	18.40	19.89	20.43
Kurtosis	−0.45	0.84	2.23	−1.24	0.97	2.15	−0.01	1.93	1.36	−0.75	2.94	−0.51
Skewness	0.90	1.03	1.56	0.10	−0.94	−1.45	1.12	1.29	−1.58	0.15	0.58	0.47

, reporting the minimum, maximum, average, median, standard deviation, kurtosis, and skewness for the neonatal mortality rate as the dependent variable, annotated by Y, for CO₂ emissions, for PM2.5 pollution, and for the independent variables. We can observe a considerable difference between the lowest and highest values of the variables and high standard deviations compared to the averages, which lead to a rather high variability in the data. The average health expenditure per capita (${\mathrm{X}}_4$ (2341.42) is almost double the median (1208.45), indicating that half of the countries spend less than USD 1208.45 per capita, and half of the countries spend more than USD 1208.45 for the health sector. Similarly, the average number of hospital beds per 1000 people (${\mathrm{X}}_5$) (32.11) is considerably higher than the median (25.02), which indicates that half of the countries have less than 25 beds, and half of the countries have more than 25 inpatient beds available in public, private, general, and specialized hospitals and rehabilitation centers, for both acute and chronic care. The kurtosis of the crop production index (${\mathrm{X}}_8$), the ratio of female to male labor force participation rate (${\mathrm{X}}_3$), and PM2.5 pollution is greater than 2 (2.94, 2.15, and 2.23, respectively), indicating a too-peaked distribution. The kurtosis is positive for CO₂ emissions (${\mathrm{X}}_1$), women, business and the law index score (${\mathrm{X}}_2$), number of hospital beds (${\mathrm{X}}_5$), and people using at least basic sanitation services (${\mathrm{X}}_6$), indicating a more peaked distribution. The skewness is negative for women, business and the law index score (${\mathrm{X}}_2$), ratio of female to male labor force participation rate (${\mathrm{X}}_3$), and people using at least basic sanitation services (${\mathrm{X}}_6$), indicating left-skewed distributions. Gender inequality index (${\mathrm{X}}_1$) and forest area (${\mathrm{X}}_7$) have a skewness value near zero, indicating a symmetric distribution.

We categorized the 88 countries into two distinct groups based on high versus low CO₂ emissions, using the median (3.18) as a threshold, and into two distinct groups based on PM2.5 pollution, using the median (17.10) as a threshold. There is a statistically significant difference between the average CO₂ emissions in the two groups of countries (test statistic = −12.33; p-value = 4.67E-17), as reported in

Table 3. Comparison of the average values of the variables for the countries with low versus high levels of CO₂ emissions.

	High (n = 44) (Mean, SD)	Low (n = 44) (Mean, SD)	p-Value
CO2	6.48 (2.65)	1.29 (0.86)	4.67E-17
PM2.5	14.28 (7.28)	26.80 (14.36)	2.632E-06
Neonatal mortality rate	3.30 (2.63)	14.92 (7.26)	7.34E-14
GII	0.15 (0.12)	0.41 (0.13)	2.67E-15
Women, business and the law index score	90.82 (10.48)	78.76 (10.69)	7.38E-07
Ratio of female to male labor force participation rate (%)	80.53 (9.62)	72.12 (18.09)	0.008305
Current health expenditure per capita, PPP	4008.43 (2490.42)	674.41 (669.84)	2.55E-11
Hospital beds	47.53 (24.73)	16.69 (12.97)	4.63E-10
People using at least basic sanitation services	98.36 (3.18)	85.38 (14.46)	5.041E-07
Forest area	357,611.7 (1,262,610)	232,768.7 (751,482.4)	0.5748
Crop production index	102.14 (16.50)	114.45 (21.22)	0.003224
Rural population	26.13 (14.53)	46.12 (20.70)	1.35E-06

. The average amount of PM2.5 pollution in the two groups of countries is also statistically significant (test statistic = −9.18; p-value = 4.274E-12), as shown in

Table 4. Comparison of the average values of the variables for the countries with low versus high levels of PM2.5 pollution.

	High (n = 44) (Mean, SD)	Low (n = 44) (Mean, SD)	p-Value
PM2.5	29.56 (12.72)	11.51 (3.21)	4.274E-12
CO2	2.24 (2.39)	5.52 (3.21)	5.949E-07
Neonatal mortality rate	14.10 (7.71)	4.12 (4.27)	1.867E-10
GII	0.38 (0.15)	0.18 (0.15)	1.248E-08
Women, business and the law index score	78.41 (10.48)	91.18 (10.27)	1.198E-07
Ratio of female to male labor force participation rate (%)	72.72 (18.30)	79.93 (9.72)	0.02428
Current health expenditure per capita, PPP	671.73 (647.82)	4011.11 (2492.57)	2.422E-11
Hospital beds	23.34 (21.98)	40.89 (25.02)	0.0007578
People using at least basic sanitation services	85.97 (13.56)	97.76 (7.11)	3.057E-06
Forest area	173,490.8 (358,376.7)	416,889.6 (1,417,089)	0.2748
Crop production index	115.88 (19.36)	100.71 (17.55)	0.0002274
Rural population	47.51 (19.38)	24.74 (14.26)	1.708E-08

. The test results also indicate that the average neonatal mortality rate is significantly different in countries with low CO₂ emissions versus in countries with high CO₂ emissions, and in countries with low vs. high PM2.5 pollution. Table 3 summarizes the average values and the standard deviations of the variables in the countries with high versus low CO₂ emissions. p-values below the significance level of 0.05 indicate that the differences between the averages in the two groups of countries are statistically significant. Only the average forest area is approximately the same in both groups of countries.

Table 4 summarizes the average values and the standard deviations of the variables in the countries with high versus low PM2.5 pollution. Similarly to the previous observations, only the average forest area is approximately the same in both groups of countries, indicated by the high p-value above the significance level of 0.05.

Table 3 and Table 4 reveal that countries with low CO₂ emissions have a higher average neonatal mortality rate and also higher average PM2.5 pollution than countries with high levels of CO₂ emissions. However, countries with low PM2.5 pollution have a lower average neonatal mortality rate and at the same time higher CO₂ emissions than countries with high levels of PM2.5 pollution.

Finally, we build four multiple regression models for these four groups of countries using the variables from Table 1 and evaluate each model. We compare the regression models in terms of the factors that exhibit a significant association with the neonatal mortality rate and select the best model that is statistically significant, with the highest proportion of variance in the neonatal mortality rate that can be explained by the independent variables.

The multiple linear regression models (1) and (2) study the neonatal mortality rate in terms of the environmental, health, and economic factors listed in Table 1 in countries with low versus high CO₂ emissions. Both models are statistically significant with p-values of 9.779E-15 and 1.807E-08, respectively, and with adjusted R2 metrics of 0.8451 and 0.6352, respectively. These proportions of variance in the neonatal mortality rate that can be explained by the independent variables are rather high as they are close to the maximum possible value of 1.

Table 5. Multiple linear regression results of models (1) and (2).

		Estimate	Std. Error	t-Value	p-Value
Model (1)	(Intercept)	13.689087	5.78178	2.368	0.02324 *
	X1	13.58281	5.348579	2.54	0.01543 *
	X2	−0.11115	0.049992	−2.223	0.03238 *
	X3	0.102815	0.035784	2.873	0.00669 **
	X4	−0.003	0.000854	−3.512	0.00119 **
	X6	−0.08338	0.031704	−2.63	0.01237 *
	X8	0.042391	0.021782	1.946	0.05926
Model (2)	(Intercept)	23.225189	4.4172839	5.258	0.00000592 ***
	X4	−0.00032	0.0001239	−2.61	0.012869 *
	X5	−0.03157	0.0102276	−3.087	0.003766 **
	X6	−0.17903	0.0459048	−3.9	0.000379 ***
	X7	−0.02911	0.0146633	−1.985	0.054410
	X9	0.036992	0.0203117	1.821	0.076455

Significance codes for the p-values: p < 0.001 ***, p < 0.01 **, p < 0.05 *.

reports the estimated coefficients of models (1) and (2), together with the standard errors, test statistic values, and the corresponding p-values.

The first model (1) contains variables from Table 1 that are statistically significant, indicating that the neonatal mortality rate (Y) in countries with low levels of CO₂ emissions is negatively affected by the women, business and the law index score (${\mathrm{X}}_2$), the amount of health expenditure per capita (${\mathrm{X}}_4$), and by the percentage of people using at least basic sanitation services (${\mathrm{X}}_6$). It is positively affected by GII (${\mathrm{X}}_1$), the ratio of female to male labor force participation (${\mathrm{X}}_3$), and the crop production index (${\mathrm{X}}_8$).

$$\mathrm{Y}=13.689+13.582811 {\mathrm{X}}_1-0.111151 {\mathrm{X}}_2+0.102815 {\mathrm{X}}_3-0.002999 {\mathrm{X}}_4-0.083384 {\mathrm{X}}_6+0.042391 {\mathrm{X}}_8$$

(1)

The second model (2) contains statistically significant variables from Table 1, indicating that the neonatal mortality rate (Y) in countries with high levels of CO₂ emissions is negatively affected by the amount of health expenditure per capita (${\mathrm{X}}_4$), the number of hospital beds (${\mathrm{X}}_5$), the percentage of people using at least basic sanitation services (${\mathrm{X}}_6$), and by the percentage of forest area (${\mathrm{X}}_7$). It is positively affected by the percentage of the rural population (${\mathrm{X}}_9$).

$$\mathrm{Y}=23.225-0.0003234 {\mathrm{X}}_4-0.0315712 {\mathrm{X}}_5-0.1790306 {\mathrm{X}}_6-0.0291052 {\mathrm{X}}_7+0.0369919 {\mathrm{X}}_9$$

(2)

We can observe that both models consider current health expenditure per capita and the percentage of people using at least basic sanitation services, while the other factors are different in the two models.

Multiple linear regression models (3) and (4) study the neonatal mortality rate in terms of the environmental, health, and economic factors listed in Table 1 in countries with low versus high PM2.5 pollution. Both models are statistically significant with p-values of 5.343E-15 and 1.842E-13, respectively, and with adjusted R2 metrics of 0.8399 and 0.8726, respectively. These proportions of variance in the neonatal mortality rate that can be explained by the independent variables are close to the maximum possible value of 1.

Table 6. Multiple linear regression results of models (3) and (4).

		Estimate	Std. Error	t-Value	p-Value
Model (3)	(Intercept)	57.4783610	6.2779013	9.156	2.93E-11 ***
	X3	−0.1022360	0.0389024	−2.628	0.01222 *
	X4	−0.0002872	0.0001571	−1.829	0.07509
	X5	−0.0356932	0.0117184	−3.046	0.00415 **
	X6	−0.4354584	0.0484479	−8.988	4.80E-11 ***
Model (4)	(Intercept)	25.58	8.12	3.151	0.003330 **
	X1	15.32	5.299	2.892	0.006540 **
	X2	−0.1447	0.05.713	−2.533	0.015956 *
	X3	0.1339	0.03714	3.607	0.000958 ***
	X4	−0.003749	0.001421	−2.639	0.012317 *
	X6	−0.1746	0.05105	−3.419	0.001610 **
	X7	−3.614E-06	1.476E-06	−2.449	0.019468 *
	X8	0.04595	0.02586	1.777	0.084318
	X9	−0.06117	0.03423	−1.787	0.082623

Significance codes for the p-values: p < 0.001 ***, p < 0.01 **, p < 0.05 *.

reports the estimated coefficients of models (3) and (4), together with the standard errors, test statistic values, and the corresponding p-values.

Model (3) contains variables from Table 1 that are statistically significant, indicating that the neonatal mortality rate (Y) in countries with low levels of PM2.5 pollution is negatively affected by the ratio of female to male labor force participation (X₃), amount of health expenditure per capita (X₄), the number of hospital beds (X₅), and the percentage of people using at least basic sanitation services (X₆).

$$\mathrm{Y}=57.478-0.102 {\mathrm{X}}_3-0.0002872 {\mathrm{X}}_4-0.036 {\mathrm{X}}_5-0.435 {\mathrm{X}}_6$$

(3)

Model (4) contains the statistically significant variables from Table 1, indicating that the neonatal mortality rate (Y) in countries with high levels of PM2.5 pollution is negatively affected by the women, business and the law index score (${\mathrm{X}}_2$), amount of health expenditure per capita (${\mathrm{X}}_4$), the percentage of people using at least basic sanitation services (${\mathrm{X}}_6$), the percentage of forest area (${\mathrm{X}}_7$), and the percentage of the rural population (${\mathrm{X}}_9$). It is positively affected by the GII (${\mathrm{X}}_1$), the ratio of female to male labor force participation (${\mathrm{X}}_3$), and the crop production index (${\mathrm{X}}_8$).

$$\mathrm{Y}=25.58+15.32 {\mathrm{X}}_1-0.1447 {\mathrm{X}}_2+0.1339 {\mathrm{X}}_3-0.003749 {\mathrm{X}}_4-0.1746 {\mathrm{X}}_6-3.614\mathrm{E}-06 {\mathrm{X}}_7+0.04595 {\mathrm{X}}_8-0.06117 {\mathrm{X}}_9$$

(4)

We can observe that both models consider the ratio of female to male labor force participation, current health expenditure per capita, and the percentage of people using at least basic sanitation services, while the other factors are different in the two models. Among the four regression models, model (4) is the best for explaining the proportions of variance in the neonatal mortality rate in countries with high levels of PM2.5 pollution with the studied factors as it has the highest adjusted ${\mathrm{R}}^2$ metric of 0.8726.

5. Discussion

The present study explores the relationship between environmental, health, and economic factors and the neonatal mortality rate as a proxy for public health in the context of air pollution [53]. By analyzing CO₂ emissions and PM2.5 pollution as the most prevalent air contaminants [26] in a wide range of geographical areas, we examine whether there is a significant difference in the neonatal mortality rate between countries with high and low levels of CO₂ emissions and between countries with high and low levels of PM2.5 pollution. CO₂ is a greenhouse gas with notable climatic effects [54], and CO₂ and PM2.5 are frequently employed as indicators of air quality [26]. Previous research by [12,14,15] reports positive associations between CO₂ emissions and health risks in the population. Besides the effects of CO₂ emissions, the influence of economic factors on health and on child mortality rates was also shown in previous studies [36,41,45].

Based on the results of the present study, CO₂ emissions and PM2.5 pollution are inversely related. We found out that countries with low CO₂ emissions and high levels of PM2.5 pollution have a higher average neonatal mortality rate. This can happen for instance due to low-GDP countries relying on agriculture and farming, with less industrial activity. This also results in naturally emitting less CO₂ compared to the energy-intensive industries and industrialized economies that depend on fossil fuels for manufacturing. However, in these regions, high levels of air pollution can arise from natural events such as wildfires or dust storms. At the same time, countries with weaker economies might often face higher neonatal mortality rates due to socioeconomic factors, inadequate healthcare, nutrition, or sanitation, or higher levels of gender inequality. The present work explores several potential factors that can contribute to high neonatal mortality rates, such as GII, crop production index, the ratio of female to male labor force participation, and the percentage of the rural population. These factors tend to have a positive effect, especially in countries with low levels of CO₂ emissions and with high levels of PM2.5 pollution. Factors that can contribute to the reduction in neonatal mortality rates include health expenditure per capita, the percentage of people using at least basic sanitation services, the women, business and the law index score, the percentage of forest area, the number of hospital beds, and the percentage of the rural population. All four regression models consider current health expenditure per capita and the percentage of people using at least basic sanitation services with a negative effect. In the regression models for countries with low CO₂ emissions and low PM2.5 pollution, the ratio of female to male labor force participation is a common factor; however, it has a positive effect in countries with low CO₂ emissions and a negative effect in countries with low PM2.5 pollution. Considering the models for countries with high CO₂ emissions and high PM2.5 pollution, the percentage of forest area and the percentage of rural population are also common factors with negative effects, except for the percentage of rural population in countries with high CO₂ emissions, which has a positive effect.

Current health expenditure per capita is a significant factor that inversely affects the neonatal mortality rate, regardless of the levels of CO₂ emissions and PM2.5 pollution. This result is in line with the findings of [49], a study that also detected a significant negative association between public healthcare expenditure and infant and neonatal mortality rates. The percentage of people using at least basic sanitation services is another significant factor that inversely affects the neonatal mortality rate in countries with either high or low CO₂ emission or PM2.5 pollution levels. Previous research works of [54] or of [55] also show that access to water and sanitation facilities affect the infant and maternal mortality rates. GII is a key factor affecting neonatal mortality rates in countries with low CO₂ emissions and high PM2.5 pollution, as increased GII levels tend to be associated with increased neonatal mortality rates. The GII is a significant indicator for assessing social factors influencing the health and mortality rates of infants and children, as identified by [47,56,57]. High ratios of female to male labor force participation rates also tend to be associated with high neonatal mortality rates in countries with low levels of CO₂ emissions and high PM2.5 pollution. The results of a previous study by [58] agree that agricultural and manual labor occupations contribute to an increasing risk of infant mortality. In countries that emit less CO₂ and have higher PM2.5 pollution, where the laws and regulations affect women’s economic opportunities positively, the neonatal mortality rates tend to be lower. Empowering women and providing better healthcare access and enhanced nutrition might help to lower neonatal mortality rates. This finding is in line with the studies of [59,60], which confirm the link between women empowerment and neonatal mortality rates. Countries that emit more CO₂ and have more forest areas and more hospital beds available in public, private, general, and specialized hospitals and rehabilitation centers tend to have lower neonatal mortality rates. A study by [61] has also reported that the availability of hospital beds contributes to healthcare accessibility, and deforestation affects environmental quality and the availability of resources. In countries with high CO₂ emissions, a higher rural population percentage is associated with a higher neonatal mortality rate. Based on the study of [46], this might imply that the degree of rurality may directly affect access to health services, which has a great impact on the neonatal mortality rate. On the other hand, in countries with high PM2.5 pollution, a higher rural population percentage is associated with a lower neonatal mortality rate.

The results of the present work help us to better understand the complex relationship between neonatal mortality and air pollution. We found several environmental, health, and economic factors that exhibit significant contributions to the neonatal mortality rate, both in positive and negative directions, such as GII, the women, business and the law index, the ratio of female to male labor force participation, the average spending on healthcare, the number of available hospital beds, and the percentage of people using at least basic sanitation services. Promoting gender equality is one of the strongest factors that can help in reducing the neonatal mortality rate in highly polluted countries. The availability of improved sanitation facilities in the households is also a strong indicator that can greatly help in reducing the neonatal mortality rate in every country, regardless of the level of air pollution in the country. The positive impacts of laws and regulations on women’s economic opportunities can also help to improve neonatal mortality rates in countries with high levels of air pollution. Increasing the labor force participation rate by providing more labor for goods and services production to women tends to have a negative impact on neonatal mortality; it tends to increase neonatal mortality rates in countries with high levels of air pollution. The effects of the amount of agricultural output and the number of people living in rural areas play a very minor role in relation to neonatal mortality, with neglectable effects. The number of available hospital beds for acute and chronic care in public, private, general, and specialized hospitals, as well as rehabilitation centers, tends to help to reduce neonatal mortality rates, mostly in countries with low PM2.5 pollution and at the same time high CO₂ emissions. However, this effect is rather small. Countries suffering from high levels of air pollution and with large amounts of land covered by natural or cultivated trees seem to benefit from forest areas in terms of neonatal mortality rate reduction, though this effect is rather small. Increasing the average spending on healthcare per person also has a slight impact on avoiding neonatal deaths regardless of the level of air pollution or CO₂ emissions in the country.

6. Conclusions

The emerging trends in the literature regarding the association of air pollution with public health suggest a complex and sometimes contradictory relationship; thus, this study aims to explore the unique findings of a negative correlation between air pollution and neonatal mortality rates across a sample of 88 countries. It is essential to explore detailed concepts of the neonatal mortality rate within the framework of air pollution. Our results emphasize the important requirement for a wide-ranging interconnection between governments, policymakers, and healthcare professionals to protect vulnerable communities. Our findings highlight the need for the ongoing observation of deforestation, the number of hospital beds, women empowerment, and the oversight of air quality to prevent health problems and foster a healthy community environment. Addressing our research question is essential for achieving SDG 3, which will contribute to creating healthier environments and, ultimately, healthier individuals.

This study has some significant advantages; however, there are some limitations that should be considered. A key limitation of our study is that not only were air quality data unavailable or incomplete for many countries but several other critical variables, including neonatal mortality rates and relevant confounders like socioeconomic status, were missing in a few settings. As a result, our final analysis was limited to data from only 88 countries in 2022. No more recent data (e.g., 2024) were accessible at the time of the analysis, which may limit the generalizability and precision of our exposure–response relationships, especially for countries not represented in our dataset. This may introduce bias and reduce the robustness of our neonatal mortality estimates associated with air pollution.