Hindawi

BioMed Research International

Volume 2022, Article ID 1181409, 19 pages
https://doi.org/10.1155/2022/1181409

Review Article
A Systematic Review of Factors Associated with Under-Five
Child Mortality

Madhav Kumar Bhusal and Shankar Prasad Khanal

Central Department of Statistics, Tribhuvan University, Kirtipur, Kathmandu, Nepal

Correspondence should be addressed to Madhav Kumar Bhusal; [email protected]

Received 12 April 2022; Revised 17 November 2022; Accepted 21 November 2022; Published 5 December 2022

Academic Editor: Ernesto Roldan-Valadez

Copyright © 2022 Madhav Kumar Bhusal and Shankar Prasad Khanal. This is an open access article distributed under the
Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited.

Background. Preventing the life of the newborn and reducing the entrenched disparity of childhood mortality across different
levels is one of the crucial public health problems, especially in underdeveloped and developing countries in the world.
Sustainable development goals (SDGs)-3.2 is aimed at terminating all preventable under-five child mortality and shrinking it to
25 per 1000 live births or lower than this by 2030. Several factors have been shown to be linked with childhood mortality.
Objective. This review is aimed at pointing out the significant determinants related to under-five child mortality by a
systematic review of the literature. Methods. EMBASE, PubMed, Scopus database, and Google Scholar search engine were used
for the systematic search of the literature. Special keywords and Boolean operators were used to point out the relevant studies
for the review. Original research articles and peer-reviewed papers published in the English language till August 10, 2022, were
included in the analysis and synthesis of the results. As per the Preferred Reporting Items for Systematic Review and Meta-
Analysis (PRISMA) guidelines, out of 299 studies identified from different sources, only 22 articles were ascertained for this
study. Eligible articles were appraised in detail, and relevant information was extracted and then integrated into the systematic
review. Results. Mother’s education, size of child at birth, age of mother at childbirth, place of residence, birth interval, sex of
child, type of birth (single or multiple), and birth order, along with other socioeconomic, maternal, child, health facility
utilization, and community level variables, were observed as important covariates of under-five mortality. Conclusion. Women’s
education and easy access to quality healthcare facilities should be the apex priority to lessen childhood mortality.

1. Introduction deaths per 1000 live birth in 1990. It declined by more than
half during the MDG period to 43 deaths per 1000 live births
Under-five mortality is a prime indicator of both sustainable in 2015; however, the target was to reduce by two-thirds
development goals (SDGs) and millennium development between 1990 and 2015. The progress of 15-year MDGs
goals (MDGs). The SDGs target 3.2 aims to end the prevent- shows a substantial decline in under-five child mortality
able death of under-five mortality and to lower at least as and improvement in maternal health. Neonatal deaths occu-
low as 25 per 1000 live birth for all countries [1]. It is consid- pied a higher proportion of under-five deaths in every region
ered one of the major indicators to evaluate the level of child of the world. Preterm birth complications, complications
health, and it illuminates the overall development in coun- during labor and delivery, and sepsis were reported as the
tries. Under-five mortality rate enables the demographic major causes of these neonatal deaths. Reduction of neonatal
assessment of the country’s population and serves as a major deaths with some effective intervention is the major chal-
indicator of the country’s socioeconomic development and lenge to curtail under-five mortality [3].
quality of life [2]. The MDG vision incepted in 2000 to fight The latest report by the United Nations Children’s Fund
against multidimensional poverty has remained a predomi- (UNICEF) has shown that under-five mortality declined by
nant development framework for the world by its termina- 61 percent globally between 1990 and 2020 as there were
tion in 2015. The global childhood mortality rate was 90 93 deaths per 1000 live births in 1990 to 37 deaths per
2 BioMed Research International

1000 live births in 2020. Despite this significant decline, the literature related to exploring the factors associated with
geographic disparity, economic variations among the coun- under-five mortality.
tries, children in fragile contexts, and variation in sex-
specific child mortality are the major issues of concern to 2.2. Study Selection. Articles selected from different data-
overcome the high under-five child mortality. Moreover, bases and search engine were managed by using the open-
the crisis of novel coronavirus (COVID-19) seems to have source citation manager software Zotero (http://www
a serious impact on the efforts and interventions made to .zotero.org). For this, all selected records were exported to
diminish childhood mortality. To cope with this adverse sit- Zotero from databases and search engine. Studies imported
uation, more acceleration and attention are required to were then merged, and duplicate studies were eliminated
maintain and strengthen life-saving interventions [4]. Dif- after confirming if they consist same authors and the same
ferent socio-economic variables, maternal and child-related research topic. After then, all articles were exported to Ray-
factors, and health service utilization factors like household yan (rayyan.ai) for further screening. In this phase, all the
wealth status; education level of the mother; older maternal records were categorized into two groups “include” and
age; use of polluting fuel; higher parity; lack of antenatal care “exclude.” The inclusion and exclusion decisions were done
(ANC) visits; lack of skilled birth attendance; use of antena- based on the title and abstract assessment of the articles.
tal iron and folic acid (IFA) supplementation; and tetanus Decisions for the very few records were made based on the
toxoid (TT) vaccination during pregnancy were found major title assessment of the article in case of the dearth of the
factors associated with childhood mortality by prior studies abstract. Rayyan is a free web-based tool that facilitates sys-
[5–10]. Moreover, some other studies have highlighted the tematic review [17]. It helps to screen each paper quickly by
role of community-level variables along with individual- allowing it to tag the paper into an inclusion or exclusion
level variables for the survival of children. These studies have category after a brief examination. All the records grouped
suggested to include community-level variables in the study in the inclusion category were then exported into Zotero,
for a more accurate estimate and useful epidemiological and then, the full text of each record was reviewed micro-
understanding [11–15]. scopically. Articles satisfying the inclusion criteria were
A proper understanding and identification of important finally decided to incorporate into the study. Any differences
determinants related to under-five child mortality are while examining the records were eliminated from the con-
extremely essential to implement healthcare policies, effec- sensus between authors.
tive allocation of healthcare resources, and to endorse appro-
priate interventions to reduce childhood mortality. This
systematic review is thus aimed at assessing the factors asso- 2.3. Inclusion and Exclusion Criteria. Some restrictions were
ciated with under-five mortality. created while selecting the papers to incorporate in the
review. Articles were included for systematic review if they
were (1) original research articles and grey literature match-
2. Methods ing the objective of the review; (2) peer-reviewed papers; (3)
published in the English language; and (4) published till
2.1. Data Sources and Search Strategy. Preferred Reporting August 10, 2022. Studies were excluded if (1) the full text
Items for Systematic Review and Meta-Analysis (PRISMA) of the paper is not available; (2) the studies were thesis; (3)
guidelines were followed to conduct this study [16]. Litera- the studies were based on factors related to medical causes
ture was searched systematically using different electronic of under-five death; (4) the studies were not related to the
databases including EMBASE, PubMed, and Scopus. In objective of this study; (5) studies did not contain adequate
order to include the grey literature like government docu- information; and (6) published using the same dataset as
ments, reports, and working papers, the Google Scholar published by other authors (s). The PRISMA flow chart for
search engine was used. These computer-based searches the selection process of the records is shown in Figure 1.
were done up to August 10, 2022. The exhaustive published
articles were retrieved from a specially designed search strat- 2.4. Data Extraction. All eligible papers were reviewed thor-
egy. To retrieve the relevant articles from the databases as oughly to extract relevant information and were recorded
per the objective of this study, well-constructed keywords into an excel sheet. The information extracted from each
or phrases were formulated. The words were connected by study was as follows: (1) first author of the study, (2) year
“OR” and “AND” operators so that these operators enable of publication, (3) type of study design, (4) major method/
to search the records in databases and search engines in an technique used for analysis, (5) study area/region/country,
exclusive manner. The search strategy used was [(‘childhood (6) major objective of the study, (7) sample size/data source,
mortality’ OR ‘under-five mortality’) AND (‘associated fac- and (8) findings (significant factors associated with under-
tors’ OR ‘risk factors’)]. The articles were searched by apply- five child mortality). MKB thoroughly extracted the data
ing the search strategy in the title search in Scopus, PubMed, from each included record. In order to reduce selection bias
and Google Scholar. In the case of EMBASE, this search and abate individual errors, SPK also verified the extracted
strategy was applied in the title or abstract search to extend information. Disagreement in the process of extracting the
the horizon of searches. This particular strategy was devel- data between authors was solved by reassessing the article
oped after multiple tests of various combinations of key- until having a common conclusion. The extracted data is
words in the database to include the majority of the shown in Table 1.
BioMed Research International 3

Identification Database Search engine

Records identified Records identified Records identified Additional records identified

through PubMed through SCOPUS through EMBASE through Google Scholar
(n =13) (n = 18) (n = 218) (n = 50)

Records after removing duplicates (n = 245)

Screening

Excluded records
Screened records (n = 57)
(n = 188)
Eligibility

Reasons for exclusion

Based on same dataset 12

Full-text papers evaluated for eligibility Related to health causes of death 6
(n = 22) Full text not available 3
Thesis 3
Not related with topic 9
Inadequate information 2
Included

Studies included for analysis

(qualitative synthesis) (n = 22)

Figure 1: PRISMA flowchart for the selection of studies.

2.5. Quality Assessment. National Institute of Health (NIH) while it was not reported in one (4.45%) cohort study. Addi-
quality assessment tools for observational cohort and tional information on quality assessment is included in the
cross-sectional studies were used for the critical appraisal supplementary file.
of each study [18]. For impartial judgment, both authors
(MKB, and SPK) independently ranked the quality of each 2.6. Synthesis of the Results. An explicit descriptive summary
article. Any discrepancies in the course of appraisal were of all the extracted records was prepared for the synthesis of
resolved by the consensus between authors. All studies the results. For this purpose, Table 1 is constructed. It con-
lucidly stated their objectives and study population and tains different components including the name of the first
had a participation rate of all eligible persons at least 50%. author, year of publication, the objective of the study, data
The subjects were selected from the same population in each source and sample size, and findings. Major findings of
study, and the determination of sample size was justified by included studies were assessed in a descriptive way from
all of them. Exposure of interest was not measured prior to the extracted information to point out the significant deter-
outcome by 21 (95.45%) cross-sectional studies whereas minants of under-five mortality.
exposure of interest was measured by one (4.55%) cohort
study, and it had a sufficient timeframe to observe an associ- 3. Results
ation between exposure and outcome variables. The relation
between different levels of exposure and outcome was exam- 3.1. Search Findings. The purpose of this study is to explore
ined by each study. Exposure measures were clearly defined the factors associated with under-five child mortality.
but not assessed more than once by all studies. The outcome Figure 1 shows the selection procedure and exclusion of
variable was clearly defined in each study, and the confound- records with specific reasons for rejection. Of the total 299
ing variables were adjusted by all of them. None of the stud- records identified from different databases and a search
ies reported outcome assessors blinded to the exposure engine, 245 remained after removing 54 duplicate records.
status of participants. The loss to follow-up after baseline Of the 245 records, 188 were discarded after evaluating the
was not applicable for 21 (95.45%) cross-sectional studies, title and abstract. Only 57 records remained for full-text
 4
Table 1: Characteristics of included studies (n = 22).

Data source and

First author Year Study design Method(s) Study area Objective(s) Factors associated with under-five child mortality
sample size
Working status of mother; working (AOR = 1:67, 95%CI = 1:34,
2:08, p value < 0.001), maternal highest level of education;
secondary or more with Ref to no education (AOR = 0:41, CI =
0:26, 0:63; p value < 0.001), previous death of siblings; yes
(AOR = 6:00, CI = 4:28, 8:40; p value < 0.001), no. of children
under age 5 at home; 3 and above (AOR = 0:11, CI = 0:08, 0:15, p
value < 0.001), TT vaccination at pregnancy; one TT with Ref to
never (AOR = 0:74, CI = 0:56, 0.96; p value = 0.023), 2 and more
To identify factors
Bangladesh DHS 2004, TT (AOR = 0:53, CI = 0:42, 0:66, p value < 0.001), delivery
Country-wide associated with
Abir [38] 2015 Cross-sectional Multilevel models 2007, and 2011 (combine) complications; any complications with Ref to none (AOR = 0:66,
(Bangladesh) mortality in children
(n = 16,722 live births) CI = 0:53, 0:84, p value = 0.001), contraceptive use; yes
under 5 years of age
(AOR = 0:33, CI = 0:27, 0:40, p value < 0.001), Mother’s age at
child birth; 30-39 with Ref to <20 (AOR = 1:64, CI = 1:01, 2:65, p
value = 0.046), birth rank and birth interval; 2nd/3rd birth rank, ≤2
years interval with Ref to 2nd/3rd birth rank, ≤2 years interval
(AOR = 2:18, CI = 1:48, 3:21, p value < 0.001), 4th birth rank, ≤2
years interval (AOR = 2:73, CI = 1:76, 4:23, p value < 0.001), birth
place and mode of delivery; home with Ref to health facilities
without caesarean (AOR = 0:58, CI = 0:41, 0:82, p value = 0.002)
Maternal age; 20-24 with Ref to 15-19 years (AOR = 0:74, p value
< 0.05), 25-29 years (AOR = 0:61, p value < 0.01), 30-34 years
(AOR = 0:63, p value < 0.05), wealth index; poorer with Ref to
richest (AOR = 0:73, p value < 0.05), place of residence; rural with
To describe the
Ref to urban (AOR = 0:79, p value < 0.001), ANC visits; no ANC
burden, trend, and Zambia DHS 2007, 2013-
Andegiorgish Multilevel Country-wide as compared to had at least one ANC visits (AOR = 3:17, p value <
2022 Cross-sectional associated factors of 14, and 2018 (combine)
[32] regression model (Zambia) 0.001), birth type; multiple birth with Ref single (AOR = 2:54, p
under-five mortality (n = 29,274 live births)
value < 0.001), size of child at birth; below average with Ref to
rate
average (AOR = 1:78, p value < 0.001), child sex; male with Ref to
female (AOR = 1:28, p value < 0.001), regions; eastern with Ref to
central (AOR = 1:52, p value < 0.01) Luapula (AOR = 1:50, p value
< 0.01) Muchinga (AOR = 1:43, p value < 0.01)
110 districts (10% random
sample of Ghana’s 2000
and 2010 National
Indirect To estimate the
Population and Housing
demographic under-five child Use of LPG in household for cooking (RR = 11:1%,
Country-wide Census), and indirect
Arku [28] 2016 Cross-sectional method and mortality and its social 95%CI = 3:0%, 18:8%) associated with lower
(Ghana) demographic methods
Bayesian spatial and environmental under-five mortality.
and Bayesian spatial
model risk factors
model was used to
estimate under-five
mortality
BioMed Research International
 Table 1: Continued.

Data source and

First author Year Study design Method(s) Study area Objective(s) Factors associated with under-five child mortality
sample size
Distance to fetch water; 31-60 min. with Ref to on the premises
(AOR = 1:076, CI = 1:030, 1:086) more than 60 min. (AOR = 1:096
, CI = 1:013, 1:199), source of drinking water; tap water with Ref to
unprotected water (AOR = 0:941, CI = 0:841, 0:986), cooking fuel;
electricity/gas with Ref to straw/animal dung (AOR = 0:920, CI =
0:811, 0:986), type of toilet; no toilet facility with Ref to toilet with
flush/pit latrine (AOR = 1:037, CI = 1:005, 1:157), other toilet type
(AOR = 0:969, CI = 0:850, 0:998), type of floor; cement with Ref
Structured To estimate the
Country-wide Ethiopian DHS 2011
BioMed Research International

to wood (AOR = 1:281, CI = 1:072, 1:392), earth/sand/dung

Ayele [22] 2015 Cross-sectional additive logistic under-five mortality
(Ethiopia) (n = 26,370 live births) (AOR = 1:345, CI = 1:044, 1:466), type of roof; corrugated iron/
regression model risk factors
metal with Ref to thatch/leaf/mud (AOR = 0:996, CI = 0:886,
0:998), mat/plastic sheet/wood (AOR = 0:998, CI = 0:979, 0:998),
type of wall; cane/trunk/bamboo with Ref to wood planks/
shingles (AOR = 0:811, CI = 0:801, 0:934), smoking habit; no
(AOR = 0:807, CI = 0:601, 0:987), region, place of residence; rural
(AOR = 1:094, CI = 1:014, 1:099), working status of mother; no
(AOR = 0:957, CI = 0:837, 0:995), sex of child; female
(AOR = 0:787, CI = 0:747, 0:971).
Sex of child; male (HR = 1:14, p value = 0.04), ethnic group; Peulh
Nouna DSS, (children
with Ref to Bwaba (HR = 1:39, p value = 0.03), religion; natural/
To investigate the born alive between Jan 1,
Penalized splines other with Ref to Muslim (HR = 1:43, p value = 0.02), type of
Country-wide (rural effect of multiple risk 1998, and Dec 31, 2001,
Becher [19] 2009 Cross-sectional and Cox birth; twin with Ref to single (HR = 1:85, p value < 0.01), age of
Burkina Faso) factors for childhood and mortality follow-up to
regression model mother at birth; young (<18 years) with Ref to middle (18-34)
mortality five years)
(HR = 1:29, p value < 0.01), distance to next health center; >10 km
(n =8,986 live births)
(HR = 1:39, p value < 0.01).
Preceding birth interval; 18-23 with Ref to <18 months (OR = 0:56,
95%CI = 0:36, 0:88, p value = 0.011), 24-29 (OR = 0:43, CI = 0:28,
0:65, p value < 0.001), 30-35 (OR = 0:35, CI = 0:23, 0:54, p value <
0.001), 36-41 (OR = 0:27, CI = 0:17, 0:43, p value < 0.001), 42-47
(OR = 0:19, CI = 0:11, 0:33, p value < 0.001), 48-53 (OR = 0:26, CI
To examine the effects
Conombo Country-wide Burkina Faso DHS-2010 = 0:14, 0:50, p value < 0.001), breastfeeding; never breastfeed with
2017 Cross-sectional Logistic regression of risk factors on
[20] (Burkina Faso) (n = 15,044 live births) Ref to ever (OR = 2:89, CI = 1:96, 4:26, p value < 0.001), type of
under-five mortality
birth; twin with Ref to single (OR = 3:75, CI = 2:78, 5:06, p value <
0.001), birth order; 2 with Ref to 1 (OR = 0:41, CI = 0:21, 0:80, p
value = 0.009), 3 with Ref to 1 (OR = 0:44, CI = 0:22, 0:91, p value =
0.026), size at birth; (OR = 1:17, CI = 1:07, 1:28, p value = 0.001);
region (p values < 0.05)
Place of residence; rural (HR = 3:61, 95%CI = 1:27, 10:32, p value <
0.05), age of mother; ≥32 with Ref to 15-20 years (HR = 0:32, CI =
To compare the
0:17, 0:60, p value < 0.01), birth order; 2-3 with Ref to 1 (HR = 2:60,
influence of
Multivariate CI = 1:03, 6:58, p value < 0.05), ≥4 (HR = 3:77, CI = 1:41, 10:09, p
geographical location Kenya DHS 2008-2009
Ettarh [33] 2012 Cross-sectional analysis (hazard Country-wide (Kenya) value < 0.05), wealth index; middle with Ref to low (HR = 0:74, CI
and key maternal (n = 16,162 live births)
ratio) = 0:59, 0:93, p value < 0.05), highest (HR = 0:77, CI = 0:58, 0:98, p
factors on under-five
value < 0.05), province/region (p value for HR < 0:05), duration of
deaths
breastfeeding; >12 months with Ref to <6 months (HR = 0:13, CI =
0:02, 0:84, p value < 0.05)
5
 6
Table 1: Continued.

Data source and

First author Year Study design Method(s) Study area Objective(s) Factors associated with under-five child mortality
sample size
Birth order and birth interval; 1st child with Ref to 2nd or 3rd child,
interval > 2 years (AHR = 1:42, p value < 0.001), 2nd or 3rd child,
interval ≤ 2 years (AHR = 1:48, p value < 0.001), 4th or higher,
interval ≤ 2 years (AHR = 1:89, p value < 0.001), household wealth
index; middle with Ref to rich (AHR = 1:42, p value = 0.001), poor
(AHR = 1:43, p value = 0.001), mother’s education; no education
Cox proportional To identify the factors Nigeria DHS 2003, 2008
Country-wide with Ref to secondary or higher (AHR = 1:19, p value = 0.032),
Ezeh [25] 2015 Cross-sectional hazard regression associated with under- and 2013 (n = 66,154 live
(Nigeria) place of residence; rural (AHR = 1:29, p value = 0.001), sex of
model five mortality births, combined sample)
child; male (AHR = 1:24, p value < 0.001), mode of delivery;
caesarean with Ref to noncaesarean (AHR = 1:74, p value = 0.001),
size of child at birth; small or very small with Ref to average or
large (AHR = 1:47, p value < 0.001), mother’s age; less than 20
with Ref to 30-39 (AHR = 1:44, p value = 0.004), 40-49
(ARH = 1:47, p value < 0.001)
Child vaccination; yes (IRR = 0:735, 95%CI = 0:647, 0:834), family
size; (IRR = 0:968, CI = 0:956, 0:980), age of mother; (IRR = 1:052,
CI = 1:047, 1:056), ANC visit; 1-3 with Ref to no (IRR = 0:841,
CI = 0:737, 0:960), 4 or above (IRR = 0:814, CI = 0:702, 0:944),
previous birth interval; 25-36 months with Ref to ≤24 months
(IRR = 0:836, CI = 0:787, 0:889), 37 and above (IRR = 0:728, CI
= 0:676, 0:783), use of contraceptive; yes (IRR = 0:885, CI =
0:814, 0:962), father’s education; secondary and above with Ref to
Two-part random-
no education (IRR = 0:695, CI = 0:594, 0:814), mother’s
effects regression To identify the
Country-wide Ethiopian DHS 2016 education; primary with Ref to no education (IRR = 0:785, CI =
Fenta [23] 2020 Cross-sectional model (negative potential risk factors
(Ethiopia) (n = 14,370 live births) 0:713, 0:864), father’s occupation; had working with Ref to no
binomial hurdle for child mortality
(IRR = 1:125, CI = 1:049, 1:206), place of delivery; private sector
model)
with Ref to home (IRR = 0:609, CI = 0:405, 0:916), type of births;
multiple with Ref to single (IRR = 1:355, CI = 1:249, 1:471), age of
mother at first birth; 17 and above with Ref to ≤16 (IRR = 0:711,
CI = 0:674, 0:750), birth order; 1-3 with Ref to first (IRR = 1:372,
CI = 1:262, 1:491), 4 and above (IRR = 1:487, CI = 1:373, 1:612),
religion; Muslim with Ref to orthodox (IRR = 1:255, CI = 1:129,
1:394), between enumeration area (level 2) (IRR = 0:526, CI =
0:474, 0:548)
BioMed Research International
 Table 1: Continued.

Data source and

First author Year Study design Method(s) Study area Objective(s) Factors associated with under-five child mortality
sample size
Previous dead child; yes (AHR = 15:97, 95%CI = 11:64, 21:92,
p value < 0.001), tetanus toxoids (TT) vaccination during
pregnancy; one TT with Ref to two or more TT (AHR = 1:54, CI
= 1:09, 2:16, p value 0.013), no TT (AHR = 2:39, CI = 1:89, 3:01,
p value < 0.001), contraceptives use; no (AHR = 2:03, CI = 1:57,
2:62, p value < 0.001), ethnicity; Madhesi with Ref to Brahmin/
Chhetri (AHR = 1:73, CI = 1:29, 2:32, p value < 0.001), mother’s
literacy level; cannot read with Ref to can read (AOR = 1:33, CI
Survey-based Cox To identify the factors
BioMed Research International

= 1:03, 1:72, p value = 0.031), mother’s occupation; agriculture

Nepal DHS (2001-2016)
Ghimire [37] 2019 Cross-sectional proportional Country-wide (Nepal) associated with under- with Ref to not working (AHR = 1:45, CI = 1:06, 1:96, p value =
(n = 16,802 live births)
hazard model five child mortality 0.018), skilled/professional (AHR = 2:15, CI = 1:40, 3:30, p value <
0.001), mother’s age; 20-29 with Ref to 40-49 years (AHR = 1:88,
CI = 1:24, 2:86, p value = 0.003), < 20 (AHR = 2:76, CI = 1:57,
4:85, p value < 0.001), birth rank and birth interval; 1st child with
Ref to 2nd/3rd birth rank, >2 years (AHR = 2:55, CI = 1:77, 3:68, p
value < 0.001), 4th/higher birth rank, interval > 2 years
(AHR = 0:36, CI = 0:24, 0:52, p value < 0.001), 4th/higher birth
rank, interval ≤ 2 years (AHR = 0:62, CI = 0:42, 0:91,
p value = 0.015)
ANC visits; yes (AOR = 0:61, 95%CI = 0:49, 0:74), education of
To assess under-five mother; elementary with Ref to no-education (AOR = 0:58, CI =
Open cohort mortality focusing on KHDSS in Kersa district, 0:49, 0:68), birth weight; normal with Ref to low (AOR = 0:78,
population-based Multilevel logistic KHDSS, eastern the trends and East Hararghe Zone, CI = 0:64, 0:95), big (AOR = 5:16, CI = 1:98, 13:47), window
Gutema [24] 2022
longitudinal regression Ethiopia associated factors Oromia region, Ethiopia presence in the house; yes (AOR = 0:80, 95%CI = 0:67, 0:95),
surveillance design based on 2008-2016 (n = 18,759 live births) occupation of mother; employed (AOR = 0:66, 95%CI = 0:48,
data in KHDSS 0:91), family size; two and/or less with Ref to more than 5
(AOR = 0:37, CI = 0:22, 0:37), 3-4 (AOR = 0:38, CI = 0:31, 0:45)
Birth order; first with Ref to 2-4 (HR = 1:22, 95%CI = 1:04, 1:43, p
value = 0.02), multiple birth (HR = 3:19, CI = 2:51, 4:05, p value <
DSS and DHS (1998-
To identify the effect 0.01), family size; 7-10 with Ref to ≤6 (HR = 0:79, CI = 0:69, 0:91,
Cross-sectional/ Cox proportional Sub-Saharan Africa 1999) of Burkina Faso
Hammer [21] 2006 of risk factors for p value < 0.01), >11 (HR = 0:66, CI = 0:58, 0:76, p value < 0.01),
cohort hazards regression (Burkina Faso) (n = 6,195 for DSS, n =
childhood mortality religion; Catholic with Ref to Muslim (HR = 0:84, CI = 0:73, 0:98,
4,957 for DHS live births)
p value = 0.03), traditional (HR = 1:30, CI = 1:09, 1:54,
p value < 0.01)
Mode of delivery; caesarean section with Ref to not caesarean
Logistic regression section (AOR = 1:449, 95%CI = 1:005, 2:089, p value < 0.05), size
model using To identify the risk of child at birth; average with Ref to small (AOR = 0:498, CI =
Country-wide Ghana DHS 2014
Iddrisu [29] 2020 Cross-sectional frequentist and factors of child 0:362, 0:684, p value < 0.05), large (AOR = 0:513, CI = 0:384,
(Ghana) (n = 5,884 live births)
Bayesian mortality 0:685, p value < 0.05), mother’s education; formal education with
framework Ref to no formal education (AOR = 0:766, CI = 0:596, 0:984, p
value < 0.05)
7
 8
Table 1: Continued.

Data source and

First author Year Study design Method(s) Study area Objective(s) Factors associated with under-five child mortality
sample size
To examine province-
Logistic regression level geographic
Preceding birth interval; <24 months with Ref to ≥24 months
and multivariate variation in under-five
Country-wide Congo DHS 2007 (AOR = 1:14, 95%CI = 1:04, 1:26), place of delivery; home with
Kandala [34] 2014 Cross-sectional Bayesian geo- mortality and
(DR of Congo) (n = 8,992 live births) Ref to hospital (AOR = 1:13, CI = 1:01, 1:27), marital status of
additive survival accounting risk factors
mother; single with Ref to married (AOR = 1:16, CI = 1:03, 1:33)
analysis of under-five
mortality
Maternal age; 26-30 with Ref to ≤20 years (OR = 1:70, 95%CI =
1:30, 2:22, p value 0.001), 31-35 (OR = 2:48, CI = 1:84, 3:33, p
value = 0.001), >35 (OR = 2:87, CI = 2:10, 3:91, p value = 0.001),
maternal age at first marriage; 20-24 with Ref to <15 years
(OR = 0:80, CI = 0:70, 0:90, p value = 0.001), ≥25 (OR = 0:70,
CI = 0:57, 0:85, p value = 0.001), use of contraception; traditional
with Ref to no method (OR = 0:69, CI = 0:51, 0:85, p value =
0.017), health seeking behavior; average with Ref to low
(OR = 0:06, CI = 0:05, 0:07, p value = 0.001), preceding birth
interval; 18-36 with Ref to <18 months (OR = 0:30, CI = 0:26, 0:34,
To determine risk p value = 0.001), >36 (OR = 0:09, CI = 0:07, 0:10, p value = 0.001),
Multiple logistic Country-wide Nigeria DHS 2008
Kayode [26] 2012 Cross-sectional factors of under-five breastfeeding; >18 with Ref to <6 months (OR = 0:43, CI = 0:35,
regression (Nigeria) (n = 28,647 live births)
mortality 0:53, p value = 0.001), birth order; 2, 3, or 4 with Ref to 1
(OR = 1:93, CI = 1:56, 2:37, p value = 0.001), birth weight; small
with Ref to normal (OR = 1:31, CI = 1:09, 1:58, p value = 0.004),
family size; >5 with Ref to 1-5 (OR = 3:54, CI = 3:07, 4:08, p
value = 0.001), type of toilet; bad toilet with Ref to good toilet
(OR = 1:77, CI = 1:46, 2:14, p value = 0.001), fuel source;
kerosene with Ref to gas (OR = 0:52, CI = 0:44, 0:63, p value =
0.001), others (OR = 0:28, CI = 0:23, 0:34, p value = 0.001), no.
of wives; more wives with Ref to one (OR = 1:47, CI = 1:30, 1:66, p
value = 0.001), type of residence; rural (OR = 1:53, CI = 1:16, 1:14,
p value = 0.002)
Sex of child; female (OR = 0:62, CI = 0:42, 0:91, p value = 0.016),
type of births; multiple (OR = 2:72, CI = 1:02, 7:23, p value =
To determine the 0.046), breastfeeding duration; 13-18 with Ref to 0-12 months
Multiple logistic Country-wide factors associated with Lesotho DHS -2009 (OR = 0:14, CI = 0:072, 0:27, p value < 0.001), 19 and above
Motsima [30] 2016 Cross-sectional
regression (Lesotho) under-five child (n = 3,999 live births) (OR = 0:02, CI = 0:0064, 0:0684, p value < 0.001), source of
mortality energy; other with Ref to electricity; (OR = 2:54, CI = 1:32, 4:85,
p value = 0.005), marital status; formerly married with Ref to
married (OR = 2:26, CI = 1:56, 4:37, p value < 0.001)
Mother’s education; educated (HR = 0:75, CI = 0:60, 0:93, p value
< 0.001), mother’s employment; working (HR = 1:25, CI = 1:00,
To examine the effect
1:06, p value < 0.10), birth spacing; 2-3 with Ref to <2 years
of socioeconomic
Cox proportional Country-wide Pakistan DHS 2017/18 (HR = 0:57, CI = 0:46, 0:71, p value < 0.001), >3 years (HR = 0:56,
Naz [39] 2021 Cross-sectional status and type of
hazards regression (Pakistan) (n = 19,190 live births) CI = 0:45, 0:69, p value < 0.001), mother’s age at first birth; ≥18
residence on under-
years (HR = 0:79, CI = 0:65, 0:95, p value < 0.001), birth size;
five mortality
average with Ref to small (HR = 0:64, CI = 0:51, 0:78, p value <
0.001)
BioMed Research International
 Table 1: Continued.

Data source and

First author Year Study design Method(s) Study area Objective(s) Factors associated with under-five child mortality
sample size
Age of mother at first birth; >18 with Ref to ≤18 years
(AHR = 0:92, 95%CI = 0:86, 0:98, p value < 0.001), sex of child;
female (AHR = 0:90, CI = 0:84, 0:96, p value < 0.001), number of
children in the house; 3-4 child with Ref to 1-2 child (AHR = 0:40,
To point out crucial
Cox proportional Country-wide Sierra Leone DHS 2013 CI = 0:34, 0:46, p value < 0.001), 5 and above (AHR = 0:35, CI =
Naz [35] 2020 Cross-sectional risk factors of under-
hazards model (Sierra Leone) (n = 24,742 live births) 0:28, 0:41, p value < 0.001), birth interval; >3 years with Ref to <2
five mortality
years (AHR = 0:70, CI = 0:49, 0:98, p value < 0.05), size of child at
birth; smaller than average with Ref to very small (AHR = 0:56,
BioMed Research International

CI = 0:47, 0:66, p value < 0.001), average or larger (AHR = 0:55,

CI = 0:46, 0:65, p value < 0.001)
To investigate the Type of residence; rural (AHR = 0:79, 95%CI = 0:67, 0:93),
trends and mother’s education; primary with Ref to secondary or higher
determinants of Tanzanian DHS 2004- (AHR = 1:38, CI = 1:06, 1:80), birth rank and birth interval; first
neonatal, 2005, 2010, 2015-2016 child with Ref to 2 or 3 child, interval > 2 (AHR = 1:39, CI =
Cox proportional Country-wide
Ogbo [36] 2019 Cross-sectional postneonatal, infant, (n = 25,951 live births, 1:18, 1:63), 2 or 3 child, interval ≤ 2 (AHR = 1:43, CI = 1:14, 1:79),
hazards model (Tanzania)
child, and under-five combine of three 4 or more child, interval ≤ 2 (AHR = 1:58, CI = 1:27, 1:98), sex of
mortalities in Tanzanian DHS) child; male (AHR = 1:21, CI = 1:07, 1:37), size of child at birth;
Tanzania from 2004 small or very small with Ref to average or larger (AHR = 1:90,
to 2016 CI = 1:59, 2:27)
To examine the effect
Education of mother (beta coefficient = 34:44, p value = 0.019),
of demographic and
main occupation of mother (beta coefficient = 274:48, p value =
Cross-sectional Binary logistic Federal capital socioeconomic Primary data (n = 200
Rhoda [27] 2019 0.005), ethnic group (beta coefficient = 617:81, p value = 0.091),
descriptive study regression territory of Nigeria characteristics of live births)
average monthly income of mother (beta coefficient = 300:42,
women to under-five
p value = 0.064)
child mortality
Type of birth; twin (CDHS 2010: AHR = 2:08, 95%CI = 1:05,
4:13), birth interval; 2-3 years with Ref to <2 years (CDHS
2010: AHR = 0:49, CI = 0:32, 0:76; CDHS 2014: AHR = 0:48,
CI = 0:24, 0:95), more than 3 years (CDHS 2010: AHR = 0:59,
CI = 0:41, 0:86; CDHS 2014: AHR = 0:47, CI = 0:25, 0:87), age of
mother at child birth; more than 40 years with Ref to <20 years
(CDHS 2010: AHR = 3:55, CI = 1:80, 7:03; CDHS 2014:
AHR = 3:21, CI = 1:13, 9:08), mother’s education; primary with
To define persistent Cambodian DHS 2010
Ref to no education (CDHS 2010: AHR = 1:41, CI = 1:04, 1:91),
and emerging factors and 2014 (n = 8232
Weibull hazards Country-wide secondary or higher (CDHS 2010: AHR = 1:86, CI = 1:16, 2:97;
Vanthy [40] 2019 Cross-sectional associated with under- Cambodian DHS 2010,
regression (Cambodia) CDHS 2014: AHR = 1:95, CI = 1:05, 3:62), place of residence;
five mortality in n = 7,165 Cambodian
rural CDHS 2014: AHR = 2:99, CI = 1:28, 6:97), region; plain
Cambodia DHS 2014)
with Ref to Phnom Penh (CDHS 2010: AHR = 2:92, CI = 1:15,
7:39), ANC visit; have ANC (CDHS 2010: AHR = 0:42, CI =
0:29, 0:62; CDHS 2014: AHR = 0:33, CI = 0:18, 0:59), TT
vaccination; received >2 dose with Ref to not received at last
birth (CDHS 2010: AHR = 0:66, CI = 0:45, 0:97), child
vaccination status; not fully immunized with Ref to fully
immunized (CDHS 2010: AHR = 01:64, CI = 1:40, 1:93; CDHS
2014: AHR = 3:90, CI = 3:13, 4:86)
9
 10

Table 1: Continued.

Data source and

First author Year Study design Method(s) Study area Objective(s) Factors associated with under-five child mortality
sample size
Age of mother at first birth; (Mali: HR = 1:07, 95%CI = 1:05, 1:09,
p value < 0.05; Zimbabwe: HR = 1:07, CI = 1:03, 1:09, p value <
0.05), place of residence; rural (Chad: HR = 1:11, CI = 1:01, 1:19;
DR Congo: HR = 1:29, CI = 1:02, 1:57; Mali: HR = 1:28, CI =
1:01, 1:64; Niger: HR = 1:14, CI = 1:01, 1:33; Zimbabwe: HR =
1:01, CI = 0:83, 1:20, p value < 0.05), education of mother;
secondary with Ref to no formal (Zimbabwe: HR = 0:62, CI =
0:38, 0:99, p value < 0.05; higher: HR = 0:47, CI = 0:23, 0:96, p
value < 0.05), wealth index; richer with Ref to poorest (DR Congo:
HR = 0:89, CI = 0:79, 0:99, p value < 0.05), richest (DR Congo:
Multicountry analysis HR = 0:78, CI = 0:65, 0:94; Niger: HR = 0:84, CI = 0:61, 0:95, p
(five sub-Saharan value < 0.05), marital status; not currently married with Ref to
To examine the
Multivariable Cox African countries- DHS data from five sub- currently married/in union (DR Congo: HR = 1:24, CI = 1:11,
maternal factors
Yaya [31] 2018 Cross-sectional proportional Chad, Demographic Saharan Africa countries 1:40; Mali: 2.43, CI = 1:63, 3:64; Niger: HR = 1:59, CI = 1:24,
associated with under-
hazards regression Republic (DR) of (n = 40,754 live births) 2:30; Zimbabwe: HR = 1:33, CI = 1:06, 1:67, p value < 0.05), type
five mortality
Congo, Mali, Niger, of birth; multiple (Niger: HR = 1:14, CI = 1:04, 1:31; Zimbabwe:
and Zimbabwe) HR = 1:19, CI = 1:01, 1:57, p value < 0.05), mode of delivery;
caesarean section (Chad: HR = 1:32, CI = 1:00, 1:77; DR Congo:
HR = 1:20, CI = 1:01, 1:43; Mali: HR = 1:42, CI = 1:08, 1:85;
Niger: HR = 1:43, CI = 1:06, 1:92; Zimbabwe: HR = 1:49, CI =
1:03, 2:15; p value < 0.05), size of child; small with Ref to large
(DR Congo: HR = 1:13, CI = 1:02, 1:19; Niger: HR = 1:15, CI =
1:02, 1:22, p value < 0.05), birth order; 5 and above (Chad:
HR = 0:44, CI = 0:40, 0:49; DR Congo: 0.44, CI = 0:39, 0:48, p
value < 0.05), birth interval; 18-24 with Ref to <18 (DR Congo:
HR = 0:85, CI = 0:74, 0:97), >24 (Chad: HR = 0:88, CI = 0:79,
0:98; DR Congo: HR = 0:85, CI = 0:75, 0:95, p value < 0.05)
OR: odds ratio; RR: risk ratio; AOR: adjusted odds ratio; CI: confidence interval; HR: hazard ratio; AHR: adjusted hazard ratio; IRR: incidence rate ratio; Ref: reference category; DHS: demographic and health
survey; DSS: demographic surveillance system; KHDSS: Kersa health and demographic surveillance site.
BioMed Research International
BioMed Research International 11

analysis. After the detailed assessment of these 57 records,

1
only 22 articles were found eligible for systematic review sat-
2
isfying the inclusion criteria. The 35 records were discarded
from the analysis for different reasons. Out of 35, 12 studies
were removed due to the reason of using the same dataset as 1
used by other studies, but one most comprehensive study
based on such dataset is included in this review. Six studies
were found concentrated to explore the causes of under-
five mortality and related to medical reasons like lower
respiratory infections, infections, preterm birth complica-
tions, nutrition, and diarrhea. These articles were removed
from this review. Likewise, nine records included in the
selection process were found irreverent to achieve the objec-
tive of this study. Such studies which only assessed the rela-
tionship between the socioeconomic status of the family and 18
under-five mortality, assessing the factors associated with
under-five mortality before and after the health campaign
made by a community group, and studies conducted to mea- Africa South-East Asia
sure under-five mortality across spatial units were excluded. Eastern Mediterranean Western Pacific
Further, two records that did not contain sufficient informa-
tion were also omitted from this review. One such discarded Figure 2: Studies included in systematic review from different
study was restricted to the factors allied to under-five mor- regions (n = 22).
tality of boys and girls separately, and another study was
limited to maternal factors only. religion, TT vaccine taken during pregnancy, and mode of
delivery were highly influential to determine the childhood
3.2. Characteristics of Included Studies. As shown in Figure 2, mortality.
22 eligible studies (listed in Table 1) were selected for system-
atic review from four different regions, namely, Africa, South- 3.4. Thematic Distribution of Factors Related to Under-Five
East Asia, Eastern Mediterranean, and Western Pacific. Child Mortality. Various factors associated with under-five
Out of 22 studies, 18 studies were conducted in 11 coun- child mortality extracted from the review are listed in the last
tries in the Africa region (Burkina Faso [19–21]: 3, Ethiopia column of Table 1. The micro-level scrutiny and analysis of
[22–24]: 3, Nigeria [25–27]: 3, Ghana [28, 29]: 2, Lesotho reviewed articles provided 47 different factors responsible
[30]: 1, sub-Saharan (combine study of five-country; Chad, for under-five child mortality. These factors are distributed
Democratic Republic (DR) of Congo, Mali, Niger, and Zim- in six distinct thematic groups, namely, socio-economic fac-
babwe) [31]: 1, Zambia [32]: 1, Kenya [33]: 1, DR Congo tors, maternal factors, utilization of healthcare related fac-
[34]: 1, Sierra Leone [35]: 1, and Tanzania [36]: 1). One tors, child related factors, and community level factors, as
study included was conducted in Nepal [37] and another shown in Figure 4. The details of thematic factors are
one in Bangladesh [38] from South-East Asian region. Only depicted in Table 2.
one article was eligible to be included in this review from Maternal and socioeconomic factors were found equally
Pakistan [39], the Eastern Mediterranean region, and the and highly important to measure under-five child mortality.
next study selected was from Cambodia [40] from the West- Out of 47, 13 (27.7%) factors were socioeconomic and
ern Pacific region. maternal factors. The next two equally 8 (17%) of the total
significant thematic factors were child related and utilization
3.3. Summary of the Results. Various significant determi- of healthcare related factors. Similarly, 3 (6.4%) and 2 (4.2%)
nants of under-five child mortality were observed from a factors out of 47 were community level and paternal factors,
comprehensive review of articles. Figure 3 shows the factors respectively.
associated with under-five child mortality observed from at
least two studies in the review. Out of 22 articles reviewed 3.4.1. Factors Associated with Under-Five Child Mortality in
for the analysis, 11 (50%) articles reported that education the Studies Conducted in the Africa Region. Figure 5 shows
of mother is a principal determinant of under-five mortality. factors identified by two or more studies conducted in Afri-
Size of child at birth, age of mother childbirth, place of res- can countries. Among 18 articles, 9 (50%) articles [20,
idence whether family resides in the rural or urban area, 24–26, 29, 31, 32, 35, 36] found that the size of child at birth
and birth space/interval are prime variables determining is a major determinant playing an important role in under-
the survival of under-five children. Sex of child, type of birth five mortality. Age of mother, place of residence, sex of child,
(singleton or multiple/twin), birth order, working status of and mother’s education are other important variables
mother, region, type of fuel for cooking, household’s wealth explored by 7 (38.9%) studies conducted in these countries.
index, family size, duration of breastfeeding, birth order and Similarly, birth order, type of birth, birth interval, type of
birth interval, ANC visits, use of contraceptive, age of fuel for cooking, household’s wealth index, family size, reli-
mother at first birth, marital status of mother, ethnic group, gion, duration of breastfeeding, marital status, religion,
12 BioMed Research International

Previous death of sibling

Child vaccination
Place of delivery
Type of toilet
Mode of delivery
TT vaccination during pregnancy
Factors associated with under-five child mortality

Religion
Ethnic group
Marital status
Age of mother at first birth
Use of contraceptive
ANC visits
Birthorder & birth interval
Duration of breastfeeding
Family size
Household's income/wealth index
Type of fuel for cooking
Region
Working status of mother
Birth order
Type of birth (S/M)
Sex of child
Birth space/interval
Place of residence
Age of mother at child birth
Size of child at birth
Mother's education
0 2 4 6 8 10 12
Number of studies

Figure 3: Number of studies and factors associated with under-five child mortality identified from studies (n = 22).

Paternal factors, 4.2%

Community level factors,
6.4%

Maternal factors, 27.7%

Utilization of health care

related factors, 17%

Child related factors, 17%

Socio-economic factors,
27.7%

Figure 4: Percentage distribution of thematic factors related to under-five child mortality (n = 47).

ANC visits, and mode of delivery are also key variables to birth order and birth interval, mother’s education, use of
measure the under-five mortality. contraceptives, TT vaccination during pregnancy, previous
death of siblings, and age of mother at childbirth as signifi-
3.4.2. Factors Related to Under-Five Child Mortality in the cant determinant are associated with under-five mortality.
Studies Conducted in the South-East Asia Region. Figure 6 The number of children under age 5 at home, ethnic group,
exhibits the maternal, child-related, health service utiliza- delivery complication, and birthplace and mode of delivery
tion, and socioeconomic covariates of under-five child mor- were also obtained as important covariates in this region.
tality in the studies conducted in Asian countries.
Out of two studies [37, 38] conducted in this region, 3.4.3. Factors Related to Under-Five Child Mortality in the
both studies explored that working status of the mother, Study Conducted in the Eastern Mediterranean Region. Only
BioMed Research International 13

Table 2: Thematic distribution of factors associated to under-five child mortality.

Thematic factors Factors associated with under-five child mortality n (%)

Marital status, age of mother at childbirth, working status of mother, age of mother
at first birth, ethnic group, religion, duration of breastfeeding, mother’s education,
Maternal factors 13 (27.7)
age at first marriage, average monthly income of mother, main occupation of
mother, delivery complication, smoking habit
Type of fuel for cooking, distance traveled for water, source of drinking water, type
of toilet, type of floor, type of roof of house, household’s income/wealth index, family
Socioeconomic factors 13 (27.7)
size, no. of children under age 5 at home, distance to health center, type of wall of
house, window presence in house, no. of wives
Type of birth, birth order, sex of child, size of child at birth, birth order and birth
Child-related factors 8 (17)
interval, previous death of sibling, birth interval, no. of children at house
Place of delivery, child vaccination, ANC visits, use of contraceptive, TT
Utilization of health care-related
vaccination during pregnancy, mode of delivery, health seeking behavior, 8 (17)
factors
birthplace and mode of delivery
Community level factors Place of residence, region, variations between enumeration areas 3 (6.4)
Paternal factors Father’s education, occupation of father 2 (4.2)

10
9
8
Number of studies

7
6
5
4
3
2
1
0
Place of residence

Mother's education

Birth space/interval
Type of fuel for cooking

Age of mother at first birth

Birth order
Type of birth (S/M)

Family size
Region

Religion

Ethnic group
Duration of breastfeeding
Marital status

ANC visits

Birthorder & birth interval

Use of contraceptive
Size of child at birth
Age of mother at child birth

Sex of child

Household's income/wealth index

Type of toilet
Mode of delivery

Place of delivery
Working status of mother

Factors associated with under-five child mortality

Figure 5: Number of studies and factors related to under-five mortality in Africa region (n = 18).

one study [39] selected in this review was conducted in Paki- tality using 2010 and 2014 Cambodian demographic and
stan, an Eastern Mediterranean country. Mother’s educa- health survey datasets. It has explored that type of birth
tion, whether she is educated or not; working status of the whether the birth was single or multiple, birth interval, age
mother, whether she is involved in some work or not; birth of mother at childbirth, mother’s education, place of resi-
interval, mother’s age at first birth and size of child at birth, dence whether the family stay in rural or urban areas, region,
whether the weight of the child at birth was average or small ANC visits, dose of TT vaccine received, and status of child
are the important factors related to under-five child mortal- vaccination whether the child was fully vaccinated or not are
ity obtained by this study. the major covariates of under-five mortality.

3.4.4. Factors Related to Under-Five Child Mortality in the 3.4.5. Common Factors Related to Under-Five Child
Study Conducted in the Western Pacific Region. One study Mortality in the Studies Conducted in African, South-East
conducted in Cambodia [40] representing from Western Asia, Eastern Mediterranean, and Western Pacific Regions.
Pacific region studied the determinants of under-five mor- Among the significant determinants of under-five child
14 BioMed Research International

2.5

Number of studies
1.5

0.5

0
Working status of mother

No. of children under age 5 at home

Mother's education

TT vaccination during pregnancy

Delivery complications
Birthorder & birth interval

Previous death of sibling

Age of mother at child birth

Ethnic group

Birth place and mode of delivery

Use of contraceptive

Factors associated with under-five child mortality

Figure 6: Number of studies and factors related to under-five mortality in South-East Asia region (n = 2).

mortality observed in this review, only one determinant has one study carried out in Cambodia from the Western Pacific
appeared common in all four regions while some other region were place of residence, region, type of birth whether
determinants appeared common only in two and three the birth was single or multiple, child vaccination, and the
regions as depicted in Figure 7. The education of the mother number of ANC visits.
was observed as a common significant covariate in all four
regions. Out of 18 studies carried out in African countries, 3.4.6. Uncommon Factors Related to Under-Five Child
7 (38.9%) studies [23–25, 27, 29, 31, 36] have reported that Mortality in the Studies Conducted in African, South-East
a mother’s education is a significant determinant of under- Asia, Eastern Mediterranean, and Western Pacific Regions.
five child mortality while both the studies conducted in In contrast to the factors that appeared common and signif-
South-East Asian countries [37, 38] and one study [39] con- icant at least in two regions as discussed above, interestingly,
ducted in each of Eastern Mediterranean and Western various factors were found uncommon and appeared signifi-
Pacific region [40] found education of mother as a signifi- cant to explain under-five child mortality in the studies
cant covariate of under-five mortality. Working status of conducted in the same regions. Figure 8 shows the several
the mother was obtained as a common covariate of under- socioeconomic, maternal, paternal, child, and community
five mortality in Africa, South-East Asia, and Eastern Medi- level determinants of childhood mortality in different studies.
terranean regions. Likewise, the variables age of the mother Sex of child, birth order, type of fuel for cooking, house-
at childbirth were found common in three regions except hold’s wealth index, family size, duration of breastfeeding,
for the Eastern Mediterranean region while birth interval marital status, religion, mode of delivery, type of toilet, and
appeared common except in the studies conducted in coun- place of delivery were obtained as significant determinants
tries of the South-East Asian region. Birth order and birth of under-five child mortality along with other factors as
interval, use of contraceptives, and ethnic group of mothers shown in Figure 8 in the studies carried out only in the Afri-
were observed as common covariates of under-five mortality can countries, whereas previous death of sibling, number of
in Africa and South-East Asian regions. TT vaccine taken children under age 5 at home, delivery complications, and
during pregnancy is another important covariate reported birthplace and mode of delivery were found important to
by the two studies conducted in South-East Asian countries explain childhood mortality only in the studies conducted
and one study conducted in a country belonging to the in South-East Asian countries [37, 38].
Western Pacific region. The size of the child at birth and
the age of mother at first birth are the next equally important 4. Discussion
variables reported by the studies conducted in African coun-
tries and one study conducted in Pakistan, belonging to the A large body of literature is available to explain the relation-
Eastern Mediterranean region. ship between the survival of under-five children and its
Similarly, other significant common factors explored by determinants around the globe. Our study examined the
the studies conducted in different African countries and results of such 22 studies selected from a systematic
BioMed Research International 15

ANC visits
Child vaccination

Factors associated with under-five child mortality

Type of birth (S/M)
Region
Place of residence
Birth space/interval
Size of child at birth
Ethnic group
Age of mother at first birth
Age of mother at child birth
TT vaccination during pregnancy
Use of contraceptive
Mother's education
Birthorder & birth interval
Working status of mother

0 2 4 6 8 10 12
Number of studies

Africa Eastern Mediterranean

South East Asia Western Pacific

Figure 7: Number of studies with common factors affecting under-five child mortality in different regions (n = 22).

Birth place and mode of delivery

Delivery complications
No. of children under age 5 at home
Previous death of sibling
No. of children in the house
No. of wives
Factors associated with under-five child mortality

Window presence in the house

Smoking habit
Type of wall of house
Main occupation of mother
Average monthly income of mother
Age at first marriage
Health seeking behavior
Occupation of father
Between enumeration areas (level 2)
Father's education
Distance to health center
Type of roof of house
Type of floor
Source of drinking water
Distance travelled for water
Place of delivery
Type of toilet
Mode of delivery
Religion
Marital status
Duration of breastfeeding
Family size
Household's income/wealth index
Type of fuel for cooking
Birth order
Sex of child

0 1 2 3 4 5 6 7 8
Number of studies

South East Asia

Africa

Figure 8: Number of studies with factors affecting the under-five child mortality separately in African and Asian countries (n = 22).
16 BioMed Research International

procedure and obtained several significant covariates associ- sibility of facing different obstetrics complications in those
ated to under-five child mortality. We have observed that mothers with short birth space as compared to those who
education of the mother, size of child at birth, age of mother have long birth intervals [47]. Type of birth, whether it was
at childbirth, place of residence, and birth interval are the a singleton or multiple, also affects the health of a newborn.
predominant factors of under-five mortality along with Different studies have revealed that multiple births have an
other covariates. Moreover, some factors appeared common inverse association with under-five mortality [19, 20, 23,
irrespective of the regions the studies were conducted, while 30–32, 40]. Such a relationship may exist due to the conse-
some covariates were not found common in the studies con- quences of many reasons. Some prevalent reasons are poor
ducted in different regions but were observed as important management of multiple births, higher possibility of birth
determinants of under-five child mortality. defects in multiple births, higher risk in pregnancy in com-
The education of mothers appeared as a key factor to parison to single birth, multiple births may cause growth
reduce under-five mortality [23–25, 27, 29, 31, 36–39]. Edu- retardation or premature birth, and other delivery complica-
cated mother seems to be highly sensible and aware of the tions [48, 49].
importance of health care utilization, nutrition, and sanita- In different studies carried out to assess the determinants
tion to improve the health of child than their counterpart. of child survival, birth order also emerged as another leading
Moreover, it is observed that even after controlling the fam- variable. The first-born child and child born with order four
ily socioeconomic status, lower maternal and paternal edu- and above (however differ from one study to another)
cation are both risk factors for under-five mortality [41, exhibit a higher risk of mortality in comparison to those in
42]. The size of the child at birth is another important risk the middle. The association between birth order and survival
factor for childhood mortality. Previous studies have shown of a child is found to be influenced by other variables, espe-
significantly less relative odds of under-five death among cially by birth spacing, age of mother, and variations in the
children whose sizes are average or above at birth as com- family [23, 49, 50]. Employment of women is one of the
pared to those whose sizes are small at birth [29, 36]. This important components of their empowerment, to make
fact implies the necessity of a balanced diet for the mother them financially independent and for the recognition of gen-
to improve the nutritional status of the child which ulti- der roles and gender relations. But many studies have shown
mately helps to have a normal size of child. The age of a higher risk of under-five mortality in those mothers who
mother at childbirth is another protective variable responsi- are employed [22, 37–39]. Such a result indicates that there
ble for the survival of the child. Poor biological and social should be feasible child-care alternatives for working women
mechanisms at young age mothers have an adverse effect instead of discouraging them to work. The regional disparity
on the health of their first child. A child born to adolescent of under-five mortality is another important issue to be
mothers exhibits fragile health outcomes and leads to a solved. Different studies found a significant relationship
higher risk of under-five death [43]. Different studies [19, between regions and under-five mortality [20, 22, 32, 33,
32, 33] have shown that younger age (generally below 20 40]. Such discrepancies in under-five mortality across the
years) of mother at childbirth revealed significantly higher region could be due to uneven access to healthcare facilities
odds of under-five mortality with reference to middle age or there might have different levels of childhood survival
(20-34) of childbearing; however, there is no particular com- programs, policies, and interventions. It is imperative to
mon age to segregate. Another prime dominant factor of explore the reasons and develop intervention strategies in
childhood mortality is the place of residence. The rural area order to reduce the gap. Many studies have revealed the
exhibited a higher risk of under-five mortality than the inverse relation between the economic status of a family
urban area in many studies [22, 25, 31, 33, 34, 40]. The and under-five mortality. Poor families are compelled to
significant disparity between rural and urban child mortality have higher risks of under-five mortality compared to rich
shows the immediate need for healthcare interventions and families [31, 39]. This disparity might exist due to several
exploration of its causes. Substantial variation in rural- reasons. Families with poor economic status become unable
urban child mortality across socio-economic, biodemo- to afford for expensive health care services in need, they may
graphic, and proximate factors was observed in a review not provide sufficient nutritional foods for mothers and may
study conducted using data from 35 (sub-Saharan) coun- be ignorant about the overall health care of child and
tries. To safeguard the survival of children, particularly in mother, etc.
rural areas, it is paramount to provide easy access and qual- Family size is another influential variable of child sur-
ity healthcare services and to strengthen maternal and child vival. The study conducted to explore the risk factors for
health programs [44]. The next significant covariate of childhood mortality in sub-Saharan Africa found lower
under-five mortality is birth space. Prior studies have shown mortality risks for those who were born in large households
that short birth intervals and child survival are inversely [21], whereas another study observed higher infant mortality
related. Women with short birth intervals possess higher in those households having a large number of children (3-5
odds of under-five mortality [45, 46]. Among various possi- and ≥6) in comparison to those having 1-2 children [51].
ble causes, this significant association may exist because a These two findings imply that rather than the number of
mother within the short birth interval (≤ to 18 months) family members, a large number of children could be a det-
could not be fit biologically for subsequent birth due to loss rimental factor in child mortality. The duration of breast-
of nutrients and blood loss during breastfeeding and preced- feeding is also a significant covariate of under-five child
ing pregnancy, respectively. Moreover, there is a higher pos- mortality. Past studies have shown that a longer period of
BioMed Research International 17

breastfeeding reduces the risk of under-five mortality [20, nomic, maternal, child-related factors allied to healthcare
26, 30, 33]. Breastfeeding yields sufficient natural nutrition utilization and community-level variables as important
to the newborn and protects them from different ailments, determinants of under-five mortality. Education of mother,
and it also enhances the immune system of children [52]. size of child at birth, age of mother at childbirth, place of
ANC is a maternal health care program offered by trained residence, and birth interval were the significant and most
health workers to pregnant women. Its main objectives are frequently observed covariates of under-five mortality.
to recognize the risk, prevention, and control pregnancy-
related diseases. It also offers health education for mothers Conflicts of Interest
and children [53]. Results from earlier studies revealed that
an increase in the number of ANC visits reduces under- There are no competing interests between authors regarding
five mortality [23, 24, 32, 40]. This finding suggests imple- the research, publication, and authorship of this article.
menting appropriate intervention programs to encourage
ANC visits in order to significantly reduce under-five mor- Authors’ Contributions
tality. The use of contraceptives is an important measure
to reduce childhood mortality. The likelihood of under-five MKB designed and conceptualized the proposal of the study,
mortality decreases with the use of contraceptives [23, 26, reviewed and developed a search strategy, executed searches,
37, 38]. Its uses increase the successive birth interval and and reviewed all articles satisfying the inclusion criteria.
contribute to increasing the survival of mothers and reduc- MKB extracted the relevant information from these articles,
ing childhood mortality [54]. performed the review, and prepared the manuscript. SPK
This study contains some limitations. Although the prin- verified the extracted records and reviewed and edited this
cipal objective of this study is to explore the exhaustive document. Both authors critically reviewed the content and
determinants of under-five mortality, the factors associated approved it for final submission.
with medical causes responsible for under-five mortality
are overlooked in this review. Identification of such factors Acknowledgments
besides socio-economic, maternal, child-related, health care
utilization, and community level variables may provide more We are very much grateful to all the researchers whose arti-
comprehensive information in the effort making to mitigate cles were used in this study and to all anonymous reviewers
under-five mortality. Furthermore, the prior studies which for their insightful comments and suggestions. The authors
were not available in the database and were not accessible would also like to acknowledge the Central Department
in the exploration through search engines are excluded from Research Committee (CDRC) members for their candid
this study. The results and interpretations made in this comments and suggestions as this study is part of a Ph.D.
review could be different if such studies were included in research work.
the analysis. Also, the possible discrepancies in the factors
explored by reviewed articles as a consequence of the appli- Supplementary Materials
cation of distinct statistical models for a particular study
design are disregarded in this study. Despite these limita- The supplementary information associated with this study
tions, this study, to the best of our understanding is an up- contain results of NIH quality assessment tools employed
to-date systematic review to identify the factors associated to evaluate the study. Each study was appraised using four-
with childhood mortality. The rigorous and meticulous teen different tools, and outcomes are summarized in a sup-
review process followed to ascertain the factors of childhood plementary table 1. (Supplementary Materials)
mortality provided the list of significant covariates of child-
hood mortality. Such findings are expected to be helpful to References
formulate effective healthcare policies and introduce inter-
ventions in order to reduce child mortality. Further, the [1] United Nations Children’s Fund, “Under-five mortality,”
results would be a valuable reference for planning new stud- 2021, https://data.unicef.org/topic/child-survival/under-five-
mortality.
ies based on primary data to explore the most promising
factors associated with under-five child mortality and to [2] Ministry of Health and New ERA, & ICF, Nepal Demographic
and Health Survey 2016, Ministry of Health, Kathmandu,
quantify their effects on childhood mortality.
Nepal, 2017, https://www.dhsprogram.com/pubs/pdf/fr336/
fr336.pdf.
5. Conclusion [3] United Nations, The millennium development goals report
2015, United Nations Publications, New York, 2015.
Reducing childhood mortality and improving maternal and [4] United Nations Children’s Fund, “Levels and Trend of Child
child health is a key universal health problem. The efforts Mortality: Report 2021,” in Estimates developed by the UN
made especially in the extension of healthcare programs Inter-agency Group for Child Mortality Estimation, Unicef,
and facilities in the last few decades showed a significant New York, 2021, https://data.unicef.org/resources/levels-and-
reduction in childhood mortality. However, the existence trends-in-child-mortality.
of remarkable variations in childhood mortality across dif- [5] P. Christian, C. P. Stewart, S. C. LeClerq et al., “Antenatal and
ferent levels still possesses a common prevailing challenge postnatal iron supplementation and childhood mortality in
in front of all nations. We have extracted different socioeco- rural Nepal: a prospective follow-up in a randomized,
18 BioMed Research International

controlled community trial,” American Journal of Epidemiol- [20] B. Conombo and J. Sawadogo, “Risk factors of infant and
ogy, vol. 170, no. 9, pp. 1127–1136, 2009. under-five mortality in Burkina Faso,” Research on Humani-
[6] K. B. Khadka, L. S. Lieberman, V. Giedraitis, L. Bhatta, and ties and Social Sciences, vol. 7, no. 13, 2017.
G. Pandey, “The socio-economic determinants of infant mor- [21] G. P. Hammer, B. Kouyate, H. Ramroth, and H. Becher, “Risk
tality in Nepal: analysis of Nepal Demographic Health Survey factors for childhood mortality in sub-Saharan Africa: a com-
2011,” BMC Pediatrics, vol. 15, no. 1, p. 152, 2015. parison of data from a demographic and health survey and
[7] R. Lamichhane, Y. Zhao, S. Paudel, and E. O. Adewuyi, “Fac- from a demographic surveillance system,” Acta Tropica,
tors associated with infant mortality in Nepal: a comparative vol. 98, no. 3, pp. 212–218, 2006.
analysis of Nepal demographic and health surveys (NDHS) [22] D. G. Ayele, T. T. Zewotir, and H. G. Mwambi, “Structured
2006 and 2011,” BMC Public Health, vol. 17, no. 1, p. 53, 2017. additive regression models with spatial correlation to estimate
[8] S. Naz, A. Page, and K. E. Agho, “Potential impacts of modifi- under-five mortality risk factors in Ethiopia,” BMC Public
able behavioral and environmental exposures on reducing bur- Health, vol. 15, no. 1, p. 268, 2015.
den of under-five mortality associated with household air [23] S. M. Fenta and H. M. Fenta, “Risk factors of child mortality in
pollution in Nepal,” Maternal and Child Health Journal, Ethiopia: application of multilevel two-part model,” PLoS One,
vol. 22, no. 1, pp. 59–70, 2018. vol. 15, no. 8, p. E0237640, 2020.
[9] Y. B. Nisar, M. J. Dibley, S. Mebrahtu, N. Paudyal, and [24] G. D. Gutema, A. Geremew, D. A. Mengistu, Y. M. Dammu,
M. Devkota, “Antenatal iron-folic acid supplementation and K. Bayu, “Trends and associated factors of under-five
reduces neonatal and under-five mortality in Nepal,” The Jour- mortality based on 2008-2016 data in Kersa health and demo-
nal of Nutrition, vol. 145, no. 8, pp. 1873–1883, 2015. graphic surveillance site, eastern Ethiopia,” INQUIRY: The
[10] C. T. Sreeramareddy, H. N. Harsha Kumar, and B. Sathian, Journal of Health Care Organization, Provision, and Financ-
“Time trends and inequalities of under-five mortality in Nepal: ing., vol. 59, p. 004695802210903, 2022.
a secondary data analysis of four demographic and health sur- [25] O. K. Ezeh, K. E. Agho, M. J. Dibley, J. J. Hall, and A. N. Page,
veys between 1996 and 2011,” PLoS One, vol. 8, no. 11, article “Risk factors for postneonatal, infant, child and under-5 mor-
e79818, 2013. tality in Nigeria: a pooled cross-sectional analysis,” BMJ Open,
[11] S. A. Adedini, C. Odimegwu, E. N. S. Imasiku, D. N. Ononok- vol. 5, no. 3, p. e006779, 2015.
pono, and L. Ibisomi, “Regional variations in infant and child [26] G. A. Kayode, V. T. Adekanmbi, and O. A. Uthman, “Risk fac-
mortality in Nigeria: a multilevel analysis,” Journal of Biosocial tors and a predictive model for under-five mortality in Nigeria:
Science, vol. 47, no. 2, pp. 165–187, 2015. evidence from Nigeria demographic and health survey,” BMC
[12] V. T. Adekanmbi, N.-B. Kandala, S. Stranges, and O. A. Uth- Pregnancy and Childbirth., vol. 12, no. 1, 2012.
man, “Contextual socioeconomic factors associated with [27] M. Rhoda, N. B. Beatrice, D. S. Panse, I. Sunday, and
childhood mortality in Nigeria: a multilevel analysis,” J Epide- Y. Stephen, “Demographic and socioeconomic factors associ-
miol Community Health, vol. 69, no. 11, pp. 1102–1108, 2015, ated with under-five mortality in Nigeria’s federal capital terri-
https://www.jstor.org/stable/44017630?seq=1. tory,” Journal of Social Sciences and Humanities, vol. 2, no. 2,
[13] A. G. Boco, Individual and community level effects on child pp. 43–50, 2019.
mortality: an analysis of 28 demographic and health surveys [28] R. E. Arku, J. E. Bennett, M. C. Castro et al., “Geographical
in sub-Saharan Africa, DHS working papers, Calverton, Mary- inequalities and social and environmental risk factors for
land, USA, 2010, https://dhsprogram.com/pubs/pdf/WP73/ under-five mortality in Ghana in 2000 and 2010: Bayesian spa-
WP73.pdf. tial analysis of census data,” PLoS Medicine, vol. 13, no. 6,
[14] S. N. Dwivedi, S. Begum, A. K. Dwivedi, and A. Pandey, “Com- p. e1002038, 2016.
munity effects on public health in India: a hierarchical model,” [29] A. K. Iddrisu, K. Tawiah, F. K. Bukari, and W. Kumi, “Fre-
Health, vol. 4, no. 8, pp. 526–536, 2012. quentist and Bayesian regression approaches for determining
[15] W. H. Mosley and L. C. Chen, “An analytical framework for risk factors of child mortality in Ghana,” Bio Med Research
the study of child survival in developing countries,” Population International., vol. 2020, 2020.
and Development Review, vol. 10, pp. 25–45, 1984, https:// [30] T. Motsima, “The risk factors associated with under-five mor-
www.jstor.org/stable/2807954?origin=crossref&seq=1. tality in Lesotho using the 2009 Lesotho demographic and
[16] M. J. Page, J. E. McKenzie, P. M. Bossuyt et al., “The PRISMA health survey,” International Journal of Medical and Health
2020 statement: an updated guideline for reporting systematic Sciences, vol. 10, no. 1, pp. 43–51, 2016.
reviews,” British Medical Journal, vol. 372, article n71, 2021. [31] S. Yaya, G. Bishwajit, F. Okonofua, and O. A. Uthman, “Under
[17] M. Ozzani, H. Hammady, Z. Fedorowich, and A. Elmagarmid, five mortality patterns and associated maternal risk factors in
“Rayyan- a web and mobile app for systematic reviews,” Sys- sub-Saharan Africa: a multi-country analysis,” PLoS One,
tematic Reviews, vol. 5, no. 1, p. 210, 2016. vol. 13, no. 10, article e0205977, 2018.
[18] L. L. Ma, Y. Y. Wang, Z. H. Yang, D. Huang, H. Weng, and [32] A. K. Andegiorgish, H. G. Woldu, M. Elhoumed, Z. Zhu, and
X. T. Zeng, “Methodological quality (risk of bias) assessment L. Zeng, “Trends of under-five mortality and associated risk
tools for primary and secondary medical studies: what are they factors in Zambia: a multi survey analysis between 2007 and
and which is better,” Military Medical Research, vol. 7, no. 1, 2018,” BMC Pediatrics., vol. 22, no. 1, p. 341, 2022.
p. 7, 2020. [33] R. R. Ettarh and J. Kimani, “Determinants of under-five
[19] H. Becher, G. Kauermann, P. Khomski, and B. Kouyate, mortality in rural and urban Kenya,” Rural and Remote
“Using penalized splines to model age- and season-of-birth- Health, vol. 12, no. 1, pp. 3–11, 2012.
dependent effects of childhood mortality risk factors in rural [34] N. B. Kandala, T. P. Mandungu, K. Mbela et al., “Child mortal-
Burkina Faso,” Biometrical Journal, vol. 51, no. 1, pp. 110– ity in the Democratic Republic of Congo: cross-sectional evi-
122, 2009. dence of the effect of geographic location and prolonged
BioMed Research International 19

conflict from a national household survey,” BMC Public [48] Y. Berelie, L. Yismaw, E. Tesfa, and M. Alene, “Risk factors for
Health, vol. 14, no. 1, p. 266, 2014. under-five mortality in Ethiopia: evidence from the 2016 Ethi-
[35] L. Naz, K. K. Patel, and I. E. Uzoma, “Crucial predicting factors opian demographic and health survey,” South African Journal
of under-five mortality in Sierra Leone,” Clinical Epidemiology of Child Health, vol. 13, no. 3, pp. 137–140, 2019.
and Global Health., vol. 8, no. 4, pp. 1121–1126, 2020. [49] S. G. Gebremichael and S. M. Fenta, “Under-five mortality and
[36] F. A. Ogbo, O. K. Ezeh, A. O. Awosemo et al., “Determinants associated risk factors in rural settings of Ethiopia: evidences
of trends in neonatal, post-neonatal, infant, child and under- from 2016 Ethiopian Demographic and Health Survey,”
five mortalities in Tanzania from 2004 to 2016,” BMC Public Advances in Public Health., vol. 2020, pp. 1–13, 2020.
Health, vol. 19, no. 1, p. 1243, 2019. [50] J. N. Hobcraft, J. W. McDonald, and S. O. Rutstein, “Demo-
[37] P. R. Ghimire, K. E. Agho, O. K. Ezeh, A. M. N. Renzaho, graphic determinants of infant and early child mortality: a
M. Dibley, and C. R. Greenow, “Under-five mortality and asso- comparative analysis,” Population Studies: A Journal of
ciated factors: evidence from the Nepal demographic and Demography, vol. 39, no. 3, pp. 363–385, 1985.
health survey (2001-2016),” International journal of environ- [51] M. Ekholuenetale, A. I. Wegborn, G. Tudeme, and A. Onikan,
mental research and public health, vol. 16, no. 7, p. 1241, 2019. “Household factors associated with infant and under-five mor-
[38] T. Abir, K. E. Agho, A. N. Page, A. H. Milton, and M. J. Dibley, tality in sub-Saharan Africa countries,” International Journal
“Risk factors for under-5 mortality: evidence from Bangladesh of Child Care and Education Policy., vol. 14, no. 1, 2020.
Demographic and Health Survey, 2004-2011,” BMJ Open, [52] K. M. Jackson and A. M. Nazar, “Breastfeeding, the immune
vol. 5, no. 8, article e006722, 2015. response, and long-term health,” Journal of American Osteo-
[39] L. Naz, K. K. Patel, and A. Dilanchiev, Are Socioeconomic Sta- path Association, vol. 106, no. 4, pp. 203–207, 2006.
tus and Type of Residence Critical Risk Factors of Under-Five [53] T. K. Tegegne, C. Chojenta, T. Getachew, R. Smith, and
Mortality in Pakistan, vol. 10, Clinical Epidemiology and D. Loxton, “Antenatal care use in Ethiopia: a spatial and mul-
Global Health, Evidence from nationally representative survey, tilevel analysis,” BMC Pregnancy and Childbirth., vol. 19, no. 1,
2021. p. 399, 2019.
[40] L. Vanthy, C. Chhorvann, H. Bunleng, and H. Sopheab, [54] J. Cleland, A. C. Agudelo, H. Peterson, J. Ross, and A. Tusi,
“Determinants of children under-five mortality in Cambodia: “Contraception and health,” Lancet, vol. 380, no. 9837,
analysis of the 2010 and 2014 demographic and health survey,” pp. 149–156, 2012.
International Archives of Public Health and Community Med-
icine, vol. 3, no. 2, 2019.
[41] L. Andriano and C. W. S. Monden, “The causal effect of mater-
nal education on child mortality: evidence from a quasi-
experiment in Malawi and Uganda,” Demography, vol. 56,
no. 5, pp. 1765–1790, 2019.
[42] M. Balaj, H. W. York, K. Sripada et al., “Parental education
and inequalities in child mortality: a global systematic
review and meta-analysis,” Lancet, vol. 398, no. 10300,
pp. 608–620, 2021.
[43] J. E. Finley, E. Ozaltin, and D. Canning, “The association of
maternal age with infant mortality, child anthropometric fail-
ure, diarrhoea and anaemia for first births: evidence from 55
low- and middle-income countries,” BMJ Open, vol. 1, no. 2,
article e000226, 2011.
[44] S. Yaya, O. A. Uthman, F. Okonofua, and G. Bishwajit,
“Decomposing the rural-urban gap in the factors of under-
five mortality in sub-Saharan Africa? Evidence from 35 coun-
tries,” BMC Public Health, vol. 19, no. 1, p. 616, 2019.
[45] J. C. Fotso, J. Cleland, B. Mberu, M. Mutua, and P. Elungata,
“Birth spacing and child mortality: an analysis of prospective
data from the Nairobi urban health and demographic surveil-
lance system,” Journal of Biosocial Science, vol. 45, no. 6,
pp. 779–798, 2013.
[46] D. M. Shifti, C. Chojenta, E. Holliday, and D. Loxton, “Effects
of short birth interval on neonatal, infant and under-five child
mortality in Ethiopia: a nationally representative observational
study using inverse probability of treatment weighting,” BMJ
Open, vol. 11, no. 8, article e047892, 2021.
[47] E. Budu, B. O. Ahinkorah, E. K. Ameyaw, A. A. Seidu,
B. Zegeye, and S. Yaya, “Does birth interval matter in under-
five mortality? Evidence from demographic and health surveys
from eight countries in West Africa,” BioMed Research Inter-
national, vol. 2021, Article ID 5516257, 10 pages, 2021.