Introduction

Newborn birth outcomes are a crucial community midwifery concern, as adverse birth events remain one of the leading causes of morbidity and mortality in newborns worldwide (World Health Organization, 2020). About 7% of newborns in developed countries and 16% in developing countries are affected by low birth weight (Lamichhane et al., 2015). About 5%-18% of births are preterm births (Seabrook et al., 2019). In addition, 37% of all births in developing countries result in neonatal death (Oghenetega, Ojengbede, & Ana, 2020). These challenges have resulted in a greater focus on ensuring healthy living and well-being for all individuals, especially pregnant women. This focus is captured in Sustainable Development Goal 3, which is a target statement put together by the United Nations aimed at motivating countries toward reducing mortality in neonates and children under 5 years old to below 25 per 1,000 live births (World Health Organization, 2020).

Pregnancy is considered to be a period when the health of a mother and her developing fetus (unborn child) is particularly vulnerable to the effects of natural and chemical catastrophes (Harville et al., 2017). Pregnancy is a unique life-changing period for both the pregnant woman and her fetus. Certain physiological, physical, and emotional changes occur during pregnancy as the woman interacts with her environment. This has drawn the attention of researchers to examine the environments in which pregnant women live in search of possible environmental pollutants that may be etiologies for adverse newborn outcomes such as preterm birth, low birth weight, and congenital defects (Casey et al., 2018).

Environmental pollution is the release of typically harmful substances into the environment (Dunlap, 2018). Possible sources of environmental pollution include coal, diesel traffic, generator use, petrol traffic, particulate matter, and crude oil extraction processes, among others (Janitz et al., 2019). Most of the world's environmental pollutants may be traceable to hydrocarbon-related compounds and activities. Fossil-fuel-related environmental pollution is directly related to the worldwide, insatiable demand for energy from natural gas and crude oil (Balise et al., 2016). The need for energy from crude oil has led to increased use of conventional, unconventional, and clandestine crude oil processing methods (Walker Whitworth et al., 2018). Clandestine crude oil processing, which refers to the illegal extraction of crude oil, is often performed using unconventional means. Clandestine oil processing and poor pipeline maintenance have been identified as causes of a noninsignificant percentage of oil spills in the roughly 70,000 crude-oil-producing communities around the world (Oghenetega, Ana, et al., 2020). The Niger Delta region in Nigeria has been cited as a region experiencing gross environmental pollution from oil spills caused by clandestine crude oil processing (Bruederle & Hodler, 2019).

Onshore crude oil spills may lead to irreparable environmental degradation and pose hazards to human and environmental health (Atubi, 2015). Nigeria is the second largest crude oil producer in Africa. Much of Nigeria's crude oil deposits are located in the Niger Delta region. Crude oil spills have had a significant impact on essential community resources in the Niger Delta. Communities in River State, especially those in Ogoniland like K-Dere and B-Dere, have been worst hit by the deleterious impact of environmental crude oil pollution (Yakubu, 2017). These communities have had their farmlands, playgrounds, rivers, and drinking water polluted by crude oil spills (Digha et al., 2017). Pregnant women and their newborns residing in these oil-polluted communities are vulnerable to the possible effects of these pollutants through ingestion, contact, and inhalation (Bove et al., 2019). Bove et al. found evidence that ingested fossil-fuel hydrocarbon pollutants are toxic to human cells and are able to reach the fetal side of the placenta in women who are pregnant. Therefore, the health impact of oil spills and the resultant environmental pollution on newborn birth outcomes require closer and more-rigorous examination.

The impact on newborn birth outcomes of crude-oil-spill-related environmental pollution has been copiously studied outside Africa in recent decades (Apergis et al., 2019). One study conducted by Janitz et al. (2019) in Oklahoma, USA, found no significant association between an area's oil-producing status and risk of birth defects. However, another study by Apergis et al. conducted in the same state found an association between the proximity of a mother's residence to crude oil fracking wells and risk of negative birth outcomes. Furthermore, Walker Whitworth et al. (2018) found a significant association between the proximity of a mother's residence to areas of unconventional crude oil exploration and preterm births in the Barnett Shale area of Texas, USA. Casey et al. (2018) found a significant decrease in preterm births that correlated with reduced crude oil processing in California, USA. Willis and Hystad (2019) found no association between city air pollutant exposure and risk of adverse birth outcomes in Portland, Oregon, USA. In addition, McKenzie et al. (2014) found a significant association between the density and proximity of natural gas wells within a 16-km radius of maternal residence and the prevalence of congenital defects in Colorado, USA. Few studies on the effects of crude-oil-related environmental pollution on newborn birth outcomes have been conducted in Nigeria. A pivotal study by Bruederle and Hodler (2019) found an association between onshore crude oil spills and an increase in neonatal death in Nigeria's Niger Delta region. The paucity of studies focusing on oil-producing regions outside the United States justifies the need for more studies on this issue in other countries. This study was designed to investigate the effect of environmental crude oil pollution on newborn birth outcomes in registered births in Rivers State, Nigeria.

Methods

Design

A retrospective cohort design based on birth records was used for this study. The area of study was Rivers State in Southern Nigeria's Niger Delta region. Rivers State is home to several crude-oil-producing communities. Moreover, onshore crude oil spills occur regularly in many of these communities, in part because of clandestine and conventional crude oil processing (Bruederle & Hodler, 2019). The most environmentally polluted communities in this state are in the Ogoniland area (Digha et al., 2017). Thus, K-Dere, about 52 km southeast of Port Harcourt City in the Gokana local government area in Ogoniland, Rivers State, was selected as the target exposed community in this study. K-Dere is a rural community with a population mostly engaged as farmers, petty traders, and artisans. It has one comprehensive primary health center (K-Dere Comprehensive Primary Health Centre [K-Dere PHC]), which serves a population of about 3,180 women of childbearing age within a 15-km radius. K-Dere PHC offers a broad spectrum of maternity services to mothers and community training services to midwifery students at the University of Port Harcourt.

Iriebe in the Obio/Akpor local government area in Rivers State was chosen as the comparison community. Iriebe lies about 25 km northeast of Port Harcourt City and about 41 km northwest of the K-Dere community. Iriebe is widely considered to be the least environmentally polluted region in Rivers State after Port Harcourt City. It is a semiurban area with a population mostly engaged as petty traders, artisans, and farmers. It has one primary health center (Iriebe Comprehensive Primary Health Centre [Iriebe PHC]), which serves about 2,205 women of childbearing age within a 5.2-km radius. Iriebe PHC offers a broad spectrum of maternity services to mothers and community training services to medical students at the University of Port Harcourt.

The birth records for the above two PHCs comprise case notes kept by clinicians that contain activity and event records maintained between childbirth and the first 6 weeks of follow-up. Information on background maternal characteristics, gestation, spontaneity of labor, mode of birth, birth complications, reported morbidity, and treatments administered are all documented in the birth records.

The research team considered K-Dere and Iriebe PHCs to be highly similar in terms of skill mix, quality of maternity care, and service administration and thus suitable for comparison. The status of K-Dere as a predominantly oil-spill-polluted community and Iriebe as a community largely unaffected by oil spill issues allowed the researchers to treat residents of the former as the exposure group and those of the latter as the comparison group.

Target Population

The target population for the study was the 3,088 recorded births in K-Dere and Iriebe PHCs (1,698 in K-Dere and 1,390 in Iriebe) covering a period of 5 years (2015-2019).

Sample Size Determination

A total sample size of 338 was used in the study. The minimum sample size was calculated using Cochran's (1977) sample size formula for studies involving proportion, mathematically stated as n_s = {[Z_1-[alpha]/2² x P(1 - P)] / d²} where n_s is the minimum sample size, Z_1-[alpha]/2 is the Type 1 error at p < 5% = 1.96, P is the prevalence of first parity preterm birth in mothers living in exposed areas relative to reference areas at 12.5% (Bruederle & Hodler, 2019), and d is precision = 0.05 (Charan & Biswas, 2013). A minimum required sample size of 169 birth records was calculated for each group. The total sample size of 338 is double the minimum sample size for each arm. The sample size of 169 birth records represented approximately 10% of the birth records for the study period in K-Dere PHC and approximately 12% of the birth records for the study period in Iriebe PHC.

Sampling Technique

A systematic sampling technique was utilized to select birth records for this study. The point of random origin "6" was generated using Microsoft Excel Statistical Package 2007 (Microsoft Inc., Redmond, WA, USA). The sampling interval was determined by dividing the total number of records in each of the facilities (1,698 in K-Dere and 1,390 in Iriebe) by 169, which generated systematic intervals of 10 and 8 for K-Dere and Iriebe, respectively. One hundred sixty-nine birth records were selected from registered childbirths at each PHC facility between January 2015 and December 2019.

Instrument for Data Collection

A data extraction sheet designed by the research team was used for data collection.

Validity of the instrument

To ensure face validity, the instrument was reviewed by five nursing and midwifery research experts in the Department of Nursing, University of Port Harcourt, Nigeria. These experts examined the instrument for font size, character, organization, brevity, and conformity with operationalized outcome variables. Identified areas of temporal ambiguity were corrected.

For content validity, the corrected instrument was resubmitted to another set of five research experts from the Maternal and Child Health Unit, Faculty of Clinical Sciences, University of Port Harcourt. To measure the suitability of the data extraction sheet items in measuring the intended outcome variables, these five experts were requested to score each item on the data extraction sheet dichotomously as relevant or not relevant. Analysis of the score data generated an item validity index (agreement between raters) score >= .8 and a content validity index score of .917. Following Polit and Beck (2012), a content validity index value > .8 was considered valid.

Reliability of the instrument

The interrater reliability was not examined because only one member of the research team was tasked with data collection using the instrument, whereas another member of the research team was tasked with confirming the correctness of the collected data.

Ethical Considerations

The protocol used in this study was reviewed for ethical quality and approved by the Research Ethics Committee, Office of Research Management and Development, University of Port Harcourt (approval number: UPH/CEREMAD/REC/MM68/020). Administrative permission was also obtained from the Rivers State Primary Health Care Management Board, which has oversight responsibilities for the comprehensive primary health centers. The birth records of participants were kept anonymous throughout the period of data collection. All collected data were protected and used only for the approved academic purpose.

Data Collection

Data were collected during the month of March 2020. Systematically selected birth records were scrutinized for birth outcome parameters in the most recent birth. Data were extracted using the data extraction sheet designed by the research team. The outcome variables extracted were gestation at birth, labor onset, mode of birth, birth status, birth weight, newborn weight at 6 weeks, and morbidity at 6 weeks.

Operational Definition of Terms

Gestation at birth was operationalized as preterm (< 37 weeks) or term (37-40 weeks). Labor onset was operationalized as spontaneous or assisted labor. Mode of birth was operationalized as spontaneous vaginal birth or assisted birth. Birth status was operationalized as live birth or stillbirth. Birth weight was operationalized as low birth weight (< 2.5 kg) or normal birth weight (>= 2.5 kg). Newborn weight at 6 weeks was operationalized as recorded weight (in kilograms) of the unclothed newborn at 6 weeks. Morbidity at 6 weeks was operationalized as the recorded clinical symptoms of a newborn during postnatal clinical examinations performed within 6 weeks after birth.

Data Analysis

Categorical, discrete, and continuous data were collected. Descriptive statistics were used to summarize nominal (categorical), interval (discrete), and ratio (continuous) data. The hypotheses were tested using the F test to compare the mean scores for discrete and continuous data between groups. Chi-square, Fisher exact, and relative risk tests were used to compare categorical frequencies between groups. All hypotheses were tested at a 5% level of significance.

Results

The exposed group (K-Dere) was similar to the comparison group (Iriebe) in terms of age, previous pregnancies, parity status, and previous births (p > .05). Nonetheless, the exposed group differed from the comparison group in terms of occupation (Fisher's exact test = 45.32, p < .001). The exposed group included fewer civil servants (1.8%) than the comparison group (12.4%). The background sociodemographic characteristics of the participants are summarized in Table 1.

Table 1. Sociodemographic Characteristics of Participants (Mothers;

A significant difference between the exposed group and the comparison group was found in terms of preterm birth incidents (16.0% vs. 7.7%, p = .018). The exposed group had a 108% higher incidence of preterm births than the comparison group (relative risk = 2.08, 95% CI [1.11, 3.89], p = .018). Nonetheless, onset of labor, mode of birth, birth status, and birth weight were similar between the two groups (p > .05). Data on newborn birth outcomes are summarized in Table 2.

Table 2. Newborn Birth Outcomes in Public Health Care Facilities in K-Dere and Iriebe (

Data on newborn morbidity at 6 weeks after birth are summarized in Table 3. Further data analysis revealed significant differences between the exposed and comparison groups with respect to newborn weight and morbidity within 6 weeks of birth. At 6 weeks after birth, the newborns in the exposed group weighed significantly less than those in the comparison group (4.64 +/- 0.82 vs. 4.85 +/- 0.92 kg, p = .032). In addition, morbidity within 6 weeks of birth was 203% more likely for newborns in the exposed group than in the comparison group (relative risk = 3.03, 95% CI [2.20, 4.19], p < .001).

Table 3. Newborn Morbidities at 6 Weeks After Birth in K-Dere and Iriebe (

Discussion

This study found that the exposed group experienced 2 times more preterm births than the comparison group. This finding suggests that residing in crude-oil-polluted areas such as K-Dere greatly increases the risk of experiencing preterm birth. Preterm births in K-Dere were 16.0%, whereas those in Iriebe were only 7.7%. This finding is supported by Walker Whitworth et al. (2018) who found a relationship between unconventional gas development activity and preterm birth levels in Texas, USA, showing that pregnant women in areas of crude oil exploration activities faced a 20% higher risk of experiencing preterm birth (p < .01). The similarity in findings between this study and Whitworth et al.'s study may be related to the similarity in study design used. Retrospective cohorts nested from recorded births were used in both Whitworth et al.'s study and this study. Furthermore, the finding of higher preterm birth risk found in this study was in line with Casey et al. (2018) who studied the effect of coal and oil power plant retirements in California, USA, on preterm births in nearby populations. They found a reduction in preterm birth risk of 5%-7% in areas where crude oil plants had been shut down and retired. The findings of Casey et al. suggest that a threshold of 5% or fewer preterm births is obtainable in the absence of crude oil and hydrocarbon activities and may increase to 7% if crude oil activities are introduced nearby (within 0-10 km of residence). Nevertheless, this finding contrasts with that of Willis and Hystad (2019) who studied the relationship between hazardous air pollutants and adverse birth outcomes in Portland, Oregon, USA, and found no consistent association between hydrocarbon-based air pollutants and preterm births. This divergence in findings may be explained by the focus in this study on crude oil environmental pollution, which includes farmland and water-source contamination and flaring. Willis and Hystad considered hydrocarbon air pollutants from traffic and industrial sources only. The difference in scope between this study and that of Willis and Hystad may account for the difference in results.

In this study, living in the crude-oil-polluted K-Dere area was found to be associated with a significantly increased risk of slower newborn growth rate, with body weight at 6 weeks after birth significantly lower in the exposed group than in the comparison group. This finding is supported by Bruederle and Hodler (2019) who found in a study on the impact of oil spills on infant mortality in Nigeria that pregnancies in crude-oil-polluted areas faced a significantly higher risk of negative developmental health effects in neonates and infants. The similarity in findings between this study and that of Bruederle and Hodler may be explained by both studies being conducted in the Niger Delta region. This finding was also supported by Atubi (2015) who found in a study on the effects of oil spillage on human health in nine Niger Delta communities in Delta State, Nigeria, that crude-oil-related pollution had negative effects on humans. Ingested hydrocarbon pollutants expressed in maternal breast milk may have been directly responsible for the observed slower infant growth. This assertion was supported by Digha et al. (2017) who noted in a study carried out in the Niger Delta in Nigeria that oil pollution has massively contaminated water sources, fishing areas, and livelihoods and may have contaminated the food chain.

The results of this study found that residing in the crude-oil-polluted area of K-Dere increased the risk of 6-week newborn morbidity by over 200%. This finding echoes that of Apergis et al. (2019) who found in a study on fracking and infant mortality in Oklahoma, USA, that residing within 0-10 km of crude oil processing plants was associated with negative infant health and survival indices. This finding was further buttressed by Bruederle and Hodler (2019) who found that residing near areas of oil exploration increased infant mortality by 3.8%, which effectively doubled the expected threshold. The similarity in findings between this study and Apergis et al.'s study may reflect that mostly singleton births were selected in each of the studies. The selection of singletons for assessment helps preserve participant homogeneity and the reproducibility of findings.

Limitations

This study was affected by one major limitation. The records-based retrospective cohort design utilized did not control for several potentially confounding variables, including maternal practice of adequate nutrition in pregnancy, maternal utilization of antenatal services, occupation, and compliance with prenatal drug prescriptions. This limitation may affect the validity of the conclusions. Thus, future studies may consider using a randomized controlled trial approach to improve the validity of results.

Conclusions

Pregnancies in environments affected by crude oil pollution face increased risks of preterm birth, slower newborn growth, and newborn morbidity within 6 weeks of birth. Stakeholders should lobby policymakers to promptly clean up crude-oil-polluted regions. In addition, crude oil wells should be relocated to locations that are at least 40 km distant from residential clusters, as supported by empirical studies such as Apergis et al. (2019) and Casey et al. (2018).

Author Contributions

Study conception and design: SON-O, DO, CE

Data collection: SON-O, CE

Data analysis and interpretation: All authors

Drafting of the article: All authors

Critical revision of the article: CE, SON-O, ISA, ENB-E

References