The Relationship and Pathways

between Maternal Education and

Child Nutritional Status
− Dissertation Prospectus

Dissertation Committee:
Anne R. Pebley, Chair
Majorie Kagawa-Singer
May C. Wang
Nancy E. Levine

By Kunchok Gyaltsen

November 3, 2008
Email: kunchok@ucla.edu

Outline
 Background
 Determinants of
malnutrition
  Central hypotheses
 Nepal
 Data
 Analysis plan
 Preliminary results
 Study strengths and
limitations
Background:

What is malnutrition?
 Includes both under-nutrition
(inadequate growth) and over-
nutrition (obesity).
 Malnutrition (undernurition) has
two constituents: protein-energy
 malnutrition (PEM) and
micronutrient deficiencies.
 Children’s protein-energy
malnutrition is generally assessed
with anthropometric measurement:
 Underweight (weight for age)
 Stunting (height for age)
 Wasting (weight for height)

Background (continued) :
Magnitude of malnutrition
problem among children in
developing world:

 32% (178 million) of children under

five years of age were estimated to
be stunted in all developing
countries (40% of stunted children
were found in Africa and Southeast
Asia) (2005).
  27% (or146 million) of children
under five years old or one out of
every four children are
underweight around the world
(78 million or more than half of
them are from South Asia) (2006).
 60 million children estimated to
have moderate acute malnutrition
and 13 million children under age 5
have severe acute malnutrition
(wasting) (2001).

Background (continued) :

Effect of malnutrition on
children’s health:
 Malnutrition is an essential direct
and indirect risk factor for
mortality among young children in
developing countries.
 Malnutrition is the most important
risk factor for disease, as
malnourished children have longer
and harsher illnesses.
 Malnutrition affects long-term
growth and an individual’s
physical capacity in adulthood.
 Malnutrition has important effects
on brain development and
cognition. The period of early
childhood is crucial for brain
development as well as physical
growth.

Determinants of
malnutrition:
 Basic determinants
 Underlying
determinants
  Immediate factors
Social factors:

Ethnicity

Religion

Marital status

Economic factors:

Household income

Agriculture land

Community factors:

Residence region/zone

Rural/urban

Districts health facilities

Districts health programs

Gov’t. health expenditure

Household factors:

Dwelling house quality

Household possession

Number of children

Child characteristics:

Age/Sex
Birth weight

Child activity level

Genetic endowment

Maternal education

Health knowledge

Child nutritional status

Death

Chronic growth failure/repeated illness

Empowerment:

Decision-making

Adequate dietary intake:

Breastfeeding

Supplement feeding

Food intake

Health service/environment:

Prenatal visits

Place of birth/birth attendants

Immunization

Drinking water/toilet

Food security:
Availability of food

Basic Factors

Outcome

Immediate Factors

Underlying Factors

Health behavior:

Hand washing

Smoking

Infectious diseases: Diarrhea/fever/cough

Central hypotheses:
 How does maternal education affect
children’s nutrition?

o Maternal education is an
important predictor of child
malnutrition in part because it is
a proxy for socioeconomic
status at the individual and
household levels.
 o Maternal education influences
child nutritional status because
it affects maternal health
knowledge, including
knowledge about child nutrition.
o Women’s empowerment or
participation in decision-making
is also be an important
mechanism through which
maternal education affects
children’s nutritional status.
o Educated mothers are also likely
to have better health
behaviors.
o More educated women are also
more likely to use health care
for themselves and their
children.

Nepal:
 Administrative system and land:
 o 75 districts
o 5 development regions
o 3 ecological zones
 Total population:
o 23 million (only 16% live in
Urban)
 Ethnicity and language:
o 103 ethnic/caste groups with 92
native languages
 Health indicators:
o Infant mortality rate: 56 per
1,000 (2000)
o Under five children’s mortality
rate: 74 per 1,000 (2005)
o Maternal mortality ratio: 740
per 100, 000 (2000)
o Life expectancy: 51 for female
and 52 for male (2007)
 Nutritional issue:
o 30% to 40% of Nepalese
population suffers from hunger
o 45% of the children are
underweight and 43% have
stunted

Data:
 2006 Nepal
Demographic Health
Survey (DHS)
 Nepal District Profile
2006 published by
Nepal Development
Information Institute
N=5,783

Children <5 years old

Drop total 563 cases because child not alive, respondent not presented, refused and child not
measured for other reasons

N=10,793

Women age 15 to 49 years age

Drop total 288 cases due to missing data on religion (n=3), birth size (n=1), times of breastfed
(n=1), drinking water and toilet (n=276), and land (n=5).

Drop 5 cases due to missing in respondent’s religious affiliation

N=5,220

Children measured height and weight

N=4,934

Children <5 years old

N=10,788

Women age 15 to 49 years age

Original Nepal DHS 2006 total sample:

N=8,707 households

N=10,793 women

Final women’s analytic sample

Final child analytic sample

Analysis Plan:
 Part 1: To investigate the
prevalence of children’s
malnutrition in Nepal
  Part 2: To explore the prevalence of
maternal education in different
sectors of the Nepalese population
 Part 3: To examine the effect of the
maternal education on child
nutritional status
 Part 4: To explore the role of
maternal health behavior in
relationship maternal education
and child nutritional status

Preliminary Result:
Nutritional measures
Percent
N
Weight for age:
Normal Weight
Under weight

53.5
 46.5

2,638
2,296
Height for age:
Normal height
Stunted

55.6
44.4

2,742
2,192
Weight for height:
Normal weight
Wasted

87.9
12.1

4,337
 597
Percent distribution of children’s nutritional status in Nepal
DHS 2006 (N=4,934)
 Cut-off for each measure is less than 2 standard
deviation units below the median
Nutritional status of children under five years in Nepal
(percent and age in months) DHS 2006 (N=4,943)
Comparison of stunted, wasted, and underweight among
children in Nepal and India based on Nepal and India DHS
data ( 2006)
Maternal education by women’s age group from Nepal
DHS 2006 (N=10,788)
Weight for Age (Z-score)
Maternal Education:
Normal weight (%)

Underweight (%)
No education
Primary
Secondary
Higher
46.4
 60.8
65.1
80.0
53.6
39.2
34.9
20.0
Percent distribution of maternal educational level by
weight for age (Z-score) among children age under five in
Nepal, 2006 (N=4,934).
χ2 p-value =0.000
Height for Age (Z-score)
Maternal Education:
Normal height (%)

Stunted (%)
No education
Primary
Secondary
Higher
 48.0
59.8
73.4
85.2
52.0
40.1
26.6
14.8
Percent distribution of maternal educational level by
height for age (Z-score) among children age under five in
Nepal, 2006 (N=4,934).
χ2 p-value =0.000
Weight for Height (Z-score)
Maternal Education:
Normal weight (%)

Wasted (%)
No education
Primary
Secondary
Higher
86.7
89.9
89.1
91.0
13.3
10.1
10.9
9.0
Percent distribution of maternal educational level by
weight for height (Z-score) among children age under five
in Nepal, 2006 (N=4,934).
χ2 p-value =0.000
Underweight
Stunting
Wasting
b (s.e)
p-value
eb
 b (s.e)
p-value
eb
b (s.e)
p-value
eb
Maternal edu:

No educ*
Primary
Secondary
Higher
Intercept

-
-.582
(.076)
-.765
(.094)
-1.529
(.146)
.143
(.036)

0.000
0.000
0.000
0.000

1.000
.558
.465
.217
-

-
-.479
(.075)
-1.092
(.101)
-1.827
(.163)
.079
(.036)

0.000
0.000
0.000
0.026

1.000
.
.619
.335
.160
-

-
-.311
(.120)
 -.221
(143)
-.431
(.205)
-1.879
(.053)

0.010
0.122
0.036
0.000

1.000
.732
.800
.650
-
*reference category s.e = standard error

Logistic regression result for underweight, stunting and wasting

by maternal educational level among Nepalese children (>5
years of age, 2006 (standard errors shown in parenthesis)
(N=4,934)

 Study Strengths:
o A large and nationally

representative sample of
Nepalese women and children
with comprehensive social,
economic, demographic, and
health information.
o The first comprehensive
study of the social
determinants of children’s
nutritional status in Nepal.
 Limitations:
o Lack of actual yearly or monthly
income at the household level.
o The Nepal DHS did not collect
any information on household
food security and women’s
knowledge on nutrition.
 o This study will be based on
cross-sectional data.
THANK YOU!
Questions and Comments?

Influence of Maternal Education on Child Health in Kenya

Abuya, A. Benta1, Kimani, K. James2 Onsomu, O. Elijah3
1. Abuya, A. Benta: MA, Ph.D(c), Doctoral Candidate, Education Theory and Policy, Pennsylvania State University, PA.
2. Kimani, K. James: MA, Ph.D, Postdoctoral Fellow, African Population and Health Research Center (APHRC), Shelter Afrique
Center, Longonot Road, PO Box 10787, 00100 GPO, Nairobi, Kenya
3. Onsomu, O. Elijah: MS, MPH, Ph.D(c), CHES, Doctoral Candidate, Department of Public Health Sciences, University of
North Carolina at Charlotte, NC.

Abstract
In 2003, child mortality rate was 115/1000 children in Kenya compared to 88/1000 average for Sub-
Saharan African countries. This study sought to determine effect of maternal education on immunization
(n=2,169) and nutritional status (n=5,949) on child’s health. Cross-sectional data, Kenya Demographic
Health Survey (KDHS)-2003 were used for data analyses. 80% of children were stunted and 49% were
immunized. After controlling for confounding, overall (model 8) children born to mothers with primary
education were 2.17 times more likely to be fully immunized compared to those whose mothers lacked
any formal education, p<0.001. For nutrition, unadjusted results, children born to mothers with primary
education were at 94% lower odds of having stunted growth compared to mothers with no primary
education, p<0.01. Overall (model 8) maternal education was not significant predictor-nutritional status.
Policy implications for child health in Kenya should focus on increasing health knowledge among women
for better child’s health outcomes.
Keywords: Child health; Maternal education; Immunization; Child nutrition; Health knowledge; Kenya
Correspondence to: Benta A. Abuya, College of Education, Pennsylvania State University, 300 Chambers Building University
Park, PA 16802. E-mail: atienoa6@gmail.com 2
Introduction
Prevention of child mortality is one of the Millennium Development Goals (MDGs) expected to be
achieved by 2015 (MDG Report, 2008). However, in 2008, the infant and child mortality rates for
sub-Saharan African countries and Kenya stands at 88 deaths per 1000 and 77 per 1000 children
born, respectively (PRB, 2008). In Kenya, the under five mortality rate is equally high standing at
115 for every 1000 children born (Kenya Demographic and Health Survey (KDHS), 2003); 70% of
the children succumbing to death before their fifth birthday, from childhood diseases that are
preventable (KDHS 2003; World Bank 2004). Child health remains a critical issue in Kenya, where
infant and child mortality is still substantially high; the country’s infant mortality rate stands at 67%.
Although polio had been declared not a threat to children in Kenya, 25 years ago, recent reports
indicate that the disease is now a real threat months (Capital News, 2009).
Child health is a significant marker of the quality of life in less developed countries (LDCs)
(Glewwe, 1999). Existing research continues to show a strong correlation between maternal
education and improved child health (Caldwell, 1979; Glewwe, 1999; Mirowsky & Ross 2003;
Schultz 2002). Increased years of schooling of mothers have been shown to have a positive effect on
child health. So important is the role of maternal education that it has a greater impact on children’s
health outcomes than fathers’ education, use of modern health services, and socioeconomic status
(Martin, Trussel, Salvail, & Shah, 1983; Young, Edmonson, & Andes, 1983). Cross country
comparisons have shown that in developing countries, there is an inverse relationship between higher
levels of education (particularly maternal education) and child deaths (Bicego, & Ahmad, 1996;
Bicego & Boerma, 1993; Boerma, Sommerfelt & Rustein , 1991; Cleland, & Van Ginnneken, 1988;
Cochrane, 1980; Hobcraft, McDonald, & Rustein, 3
1984; United Nations, 1985; Ware, 1984), a sign that education is still an important factor in the fight
against infant and child mortality in developing countries.
Despite the consensus that maternal education is associated with better health outcomes in children,
caveats still remain in the causal effect of maternal education on child health, and on the potential
pathways linking the two variables (Desai & Alva, 1998). Additional research is still needed to
unravel the links. The current study highlights the role of mother’s education on two key child health
outcomes—immunization and nutritional status, (measured by complete vaccination and height for
age, respectively). Immunization protects children against five childhood diseases namely: pertussis
(whooping cough); neonatal tetanus, contracted through contamination of umbilical cord at birth;
polio (a major course for lameness in the developing world); tuberculosis, which can be especially
severe in young children; and diphtheria, which is less common but kills 10%-15% of its victims
(Caldwell, 1986). Height for age is a measure of a child’s linear age—used as a measure for stunting
(KDHS, 2003). According to the KDHS (2003), stunting is an outcome of lack of adequate nutrition
for a child that usually occurs over a long period of time—representing long-term malnutrition
effects on a population of children.
In Kenya, national estimates show that among children under five years, 30% are stunted, while 11%
suffer from severe stunting. Stunting is highest among the children who are between 12-23 months.
Additionally, male children (33%) are more likely to suffer from stunting than female children (28%)
(KDHS, 2003).
Pathways linking maternal education and child health
Socio-economic status
Socio-economic status (SES) is one of the important and most researched links between mothers’
education and child health (Caldwell, 1994; Cleland & Kaufman 1993; Frost, Forste, & 4
Haas, 2005; World Bank, 1993). Education is related to the socioeconomic status of a family.
Education and socioeconomic status predict children’s health status because they influence the
individual mothers’ knowledge, attitudes and behavior, which in turn impact the health outcomes of
their children (Cleland, 1990). Caldwell & Caldwell (1993) highlight two potential mechanisms
through which education of the mother improves the health of children—through enhanced use of
health services, and, promotion of healthy behaviors and practices. Previous research attributes better
child health to the link between the educational attainment of a child’s mother and her ability to
purchase goods and services that are essential for improving the child’s health outcomes (Cleland &
Van Ginneken, 1988; Defo, 1997, Victoria, Smith & Vaughan, 1986). In this regard, education seems
to be connected to better income (Frost, Forste, & Haas, 2005), which then leads to better health
status. With increased levels of education, women are more likely to obtain steady and better paying
jobs, thereby making them more able to supplement the family income (Barrette & Brown, 1996).
Additionally, women with higher levels of education are more likely to marry better educated men,
with well paying jobs (Barrette & Brown, 1996; Cleland & Van Ginneken, 1988). Additional family
income is equated to access to better healthcare for children, including improved housing conditions
such as availability of better latrine facilities, piped water, floors that are non-dirt, and electricity.
Living in these conditions means that these households will be less contaminated on average
(Barrette & Brown, 1996: Frost, Forste, & Haas, 2005; Martin et al., 1983).
Desai & Alva (1998) found that although maternal education had a strong correlation with markers of
child health; this relationship is greatly reduced by the introduction of individual SES variables and
community effects. They conclude that maternal education acts as a proxy for SES and geographic
residence. Frost, Forste, & Haas (2005) in their Bolivian study, establish 5
that SES is the most important mechanism that links maternal education and child nutrition. Previous
research has consistently linked maternal education and SES to child health, by asserting that SES
explains a half of the effect of a mother’s education on the health outcomes for children (Cleland &
Van Ginneken, 1988; Desai & Alva, 1998).
Knowledge
School attendance by women enables them to acquire formal education, which makes them
knowledgeable on a whole range of health issues about themselves and their children. This
knowledge is imparted through a number of ways, including learning about the causes of diseases
and illnesses, recognition, prevention and curative measures, the nutritional requirements for
effective growth and development, and exposure to and synthesizing health-related messages and
recommendations from various sources, including the mass media (Caldwell, 1979; Casterline, 2001;
Castro & Juarez, 1995; Cleland, 1990; Cleland & Van Ginneken, 1988; Defo, 1997; Frenzen &
Hogan, 1982; Streatfield, Singarimbun, & Diamond, 1990).
Despite the fact that family members may contribute to how a child is raised, mothers’ knowledge is
more important because it determines how the children’s illnesses are treated (Heaton, Forste,
Hoffmann, & Flake, 2005). In developing countries, health outcomes are perceived to be influenced
by traditional beliefs about causes and symptoms of illnesses and diseases—especially in an
atmosphere of limited biomedical knowledge—where mothers will often attribute children’s illnesses
to various folk beliefs (Goldman, Pebley, & Gragnolati, 2002). On the contrary, mothers who are
more educated and with favorable views toward modern health care are expected to seek health care
assistance from medical professionals (Heaton et al., 2005). However, the evidence linking maternal
education, health knowledge and child health is not 6
conclusive. Glewwe (1999) found that health knowledge intervenes between maternal education and
child health; whereas other research in the same area has found little and sometimes no association
between maternal education and health knowledge (Cleland & Van Ginneken, 1988; Cleland 1990).
Attitudes
Education also helps to shift negative attitudes toward modern medicine by promoting awareness and
acceptance of modern medical practices and disease interpretations that are rooted in scientific data
as opposed to fatalistic beliefs (Barrette & Brown, 1996; Caldwell & Caldwell 1993; Cleland & Van
Ginneken, 1988; Defo 1997). Therefore, it is expected that educated mothers will readily embrace
modern medicine, will be more open to using preventive health care, and will not think that the
reason their child is sick is because of fate or supernatural causes (Bicego & Boerma 1993; Heaton et
al., 2005). An attitudinal link in the study of child health has been established by several studies. For
example, Zeiltin, Ghassemi & Mansour (1990) have established that mothers who are optimistic and
enterprising achieve more success with their child’s nutrition despite being poor, while children of
mothers who have a fatalistic attitudes and views often lack proper nutrition.
Autonomy
Research shows that education is linked to child health through its influence on the ability of women
to make decisions within their families (Frost, Forste, & Haas, 2005), which in turn determines how
much power she wields in the marital relationship (Heaton et al., 2005). Women usually are the
primary caregivers in the households, and in most cases, the first person in the home to recognize that
a child is sick is usually the mother (Caldwell & Caldwell, 1993; Caldwell, 1993; Frost, Forste, &
Haas, 2005). When a woman is uneducated they will often wait 7
until other household members have noticed that the child is sick—usually the persons who have
authority in the household (Caldwell et al., 1990). The more educated a woman is the more she is
able to make the initial decisions that are related to the health of her children (Jejeebhoy, 1995;
Mason, 1984).
Education increases a mother’s personal responsibility towards her children, drawing more attention
towards the child’s illness. Education demands that action should be taken and results into a visit to
the medical practitioners for the sick child instead of waiting for the decisions to be made by
traditional authority figures (Caldwell, 1979; 1993; Caldwell et al., 1990; Heaton et al., 2005).
Saraswathi (1992) further reports that improved nutritional status for female infants is dependent on
their mother’s control over family’s income. Kishor (1995) and Jejeebhoy (1995) also document that
there is a positive relationship between a mother’s autonomy and the survival of her children.
Reproductive Factors
Previous research links a mother’s reproductive factors to child health. The argument is that better
educated mothers have more control over their reproductive choices/decisions, including the number
and the spacing of births. These factors ultimately influence child health (Cleland & Van Ginneken,
1988; Mason, 1984; LeVine et al., 1994).
Research documents that in the developing countries women get married at early ages, and as a
consequence, they enter the childbearing period when their bodies are still not mature enough to
carry babies. The teenage mothers often do not use prenatal care even when these services are
available. Consequently, these young mothers are exposed to a myriad of health risks, such as
miscarriages and still births and low birth weight (LBW) (Bachrach, Clogg, & Carver 1993; Shawky
& Milaat, 2001), 8
Birth interval is an important reproductive behavior that influences the child health and survival
(Forste, 1994; Lindstrom, 2000; Manda, 1999). Three mechanisms help to explain the birth interval
and child health linkage: women who give birth in a relatively short time do not give themselves
enough time to regain their nutritional status; children born close together are often competing for
mother’s care and resources; short birth interval exposes the younger child to diseases due to
inadequate care arising from less attention by the mother (Bicego & Boerma 1993). Thus, increased
child mortality is a consequence of birth intervals that are 2 years or less apart (Sullivan et al., 1994;
Tagoe-Darko, 1995; Curtis & Steele, 1996; Gubbaju, 1986). On the other hand, children whose birth
spacing’s are is two years or more apart have a higher chance of being better cared for, are breastfed
and have better health outcomes such as proper physical growth and body weight (Bastien, 1992;
Gubbaju, 1986). In addition, Birth order/parity also influences child health and survival (Alam,
2000). However, some scholars argue that the effect of maternal education on child mortality has
nothing to do with the shifts in the reproductive behavior (Cleland & Van Ginneken, 1988). Other
scholars argue that there are reproductive linkages between mother’s education and the ability of the
child to survive (Mason, 1984).
Conceptual model
This study will be guided by the human capital and status attainment model of schooling. In the
human capital and status attainment model of schooling, it is hypothesized that schooling enables
people to acquire skills which they use to work in the various sectors of the economy (Becker 1964;
Hyman, Wright, & Reed, 1976; Sewell & Hauser, 1976; Spaeth, 1976). More years of schooling
increase people’s stock of human capital (Mirowsky & Ross, 2003). According to Mirowsky & Ross
(2003, p. 26), education is ―a learned effectiveness‖ whereby people are enabled to acquire skills,
abilities and accumulated resources which help them to 9
shape their health and well being. Therefore, education leads to an individual’s increased control over
his/her life course events, through developing the ability of gathering and interpreting information
(Mirowsky & Ross, 1989). Thus, personal control is an outcome of high levels of education and an
important pathway between education and health (Ross & Mirowsky, 1999).
Glewwe (1999), in his study of Moroccan schools connects schooling of parents to child health
through influencing and enhancing parental values, cognitive skills of parents (literacy, numeracy),
raising household income, and parents’ health knowledge. He posits that the linkage between
mother’s education and child health operate through mechanisms such as: health knowledge which is
acquired through formal education—by directly teaching health related information; literacy and
numeracy skills acquired through formal education, which help mothers in the diagnosis and
treatment of children health problems. In addition, formal schooling exposes women to modern
society, making them more receptive to modern medicine. Other studies have reported that education
acts as a source of knowledge by providing literacy, information, and cognitive skills, as well as
transforming people’s attitudes by encourages the acceptance of modern ideas that enable people to
relegate the traditional beliefs and authority ( Cleland, Bicego, & Fegan, 1991; Castro & Juarez, 1995).
The purpose of this study is to highlight the importance of mother’s education on child health in
Kenya. We seek to answer the following questions: 1) Does mother’s education affect the
immunization status of the children in Kenya? and 2) Does mother’s education affect children’s
height for age through changing nutritional status of children in Kenya?
Data and Methods
The data source for this study was the 2003 KDHS. The 2003 KDHS is the first in a series of the
DHS surveys to cover all parts of the country, including marginal areas that were 10
not previously surveyed—Turkana, Samburu, Isiolo, Marsabit, and Moyale. The KDHS 2003
collected demographic and health issues on a sample of women aged 15-49 years, and from men
aged 15-54 years. A sample selection allowed for estimation of key indicators for each of the eight
provinces, and at the same time estimate rural-urban differences. Data collection has previously been
described somewhere else (CBS, 2004). All analyses utilized the KDHS 2003 women dataset.
Measurement
This analysis focuses on the effect of mother’s education on child health, while establishing the
mediating effect of socioeconomic status (SES), access to information, attitudes, autonomy, and
mother’s reproductive variables. The covariates include partner’s education, region of residence,
mother’s age, and rural urban differences. The dependent variables are immunization status and
height for age. Child immunization information is collected for children born 35 months preceding
the 2003 KDHS. For the immunization analysis, the sample was restricted to children who were
between 12 and 35 months at the time of the survey (n=2,169). This restriction is based on the
premise that children less than 12 months may not be fully immunized and according to the World
Health Organization (WHO) recommendations, child vaccination coverage should be assessed for
children who are over the age of 12 months (Bronte-Tinkew & Dejong, 2005). The 35-month
threshold is included in the analysis to cover the cases where children were late in getting all the
eight vaccines as stipulated by the 24 month immunization schedule which is common in developing
countries.
Immunization is operationalized as the number of vaccinations a child receives. To be fully
immunized, a child should have received the following eight vaccinations: one dose of Bacille
Galvette-Guerin (BCG) vaccine, one dose of measles vaccine, three doses of polio 11
vaccine, and three doses of Diphtheria-pertusis-tetanus (DPT) vaccine (Bronte-Tinkew & Dejong,
2005). The 2003 KDHS contains data on all the eight forms of childhood immunizations. Data on the
immunization cards—both ―vaccine marked on card‖ and ―vaccine date on card‖, as well as
vaccine ―reported by mother‖ (those children whose mother reported that they were immunized but
the vaccination card was missing) were used in determining receipt of vaccinations. The cases from
the latter category (vaccination card missing) were not dropped from the analysis. Immunization
status was coded as a dummy variable, where ―1‖ means that a child has received all the eight
vaccines and ―0‖ if otherwise.
Nutritional status is measured by a child’s height for age, n=5,949 were retained for analysis. Child
height for age is a dichotomous variable, coded as ―1‖ for children who are below negative two
standard deviations of the median population and ―0‖ otherwise (Heaton et al., 2005). Height for age
is an anthropometric index that shows the growth of a child during the pre- and post-natal period. It
denotes the long term deficiencies and effects of malnutrition on health (Gillespie & Haddad, 2001).
The National Center for Health Statistics and the WHO growth reference classify children who are
below two standard deviations on the height-for-age growth curve to be stunted (Dibley, Goldsby,
Strehling, & Trowbridge, 1987).
Maternal education is the main predictor variable. The three education categories (no education,
primary, and secondary and higher) were coded as ―0‖, ―1‖, and ―2‖ respectively. Socioeconomic
status (SES) is measured by the wealth index variable, which indicates the poverty level in a
household. Wealth index thus gives one consistent measure of SES without having to build an index
out of the household ownership and household environment—which varies across different studies.
Wealth index has three categories – poor, middle, and rich coded as ―0‖, ―1‖, and ―2‖ respectively.
12
Access to information is measured by three variables: Listening to radio, newspaper reading and
watching television. Each of the variables is coded as ―0‖ does not listen to radio, reads no
newspaper, and watches no television, and ―1‖ listens to the radio, reads the newspaper and watches
television. Direct measures of knowledge, attitude, and autonomy are limited in the KDHS 2003.
Several measures are used as proxies to capture mothers’ knowledge, attitude, and autonomy.
Mothers’ knowledge is an index computed using the following variables: whether the mother has
ever heard of oral rehydration therapy for the treatment of diarrhea (ORS), whether the mothers
received AIDS information at antenatal visit and whether the mothers recognized signs of illnesses in
their children (coughing, crying, diarrhea, fever/shivering, not able to drink, and repeated vomiting).
The index for knowledge ranges from (0-7), where ―0‖ denotes no health knowledge while and ―7‖
indicates that a mother is highly knowledgeable.
Mothers’ attitude is influenced by her education and consequently learns to challenge the traditional
attitudes and beliefs making it possible for her to utilize modern healthcare services (Frost, Forste, &
Haas, 2005). According to Frost, Forste, & Haas (2005), a measure of attitude assumes that the use of
preventative health services by mothers is a sign of how receptive they are to modern health care
compared to those mothers who do not frequent emergency and curative heath care services.
Therefore, we base our measure of attitude towards on how well mothers are receptive to, and use of
preventative healthcare services. Thus, our measure for utilization of healthcare (a proxy indicator for
mothers’ attitude toward modern medicine) combines measures of receipt of prenatal care,
doctor/nurse attendant at birth, receipt of tetanus injection before birth, and use of any modern
method of contraception. The attitude index has a range of (0-6); ―0‖ denoting no receptivity to
modern healthcare use, and ―6‖ denoting high receptivity to use of modern healthcare. The KDHS
2003 lacks measures of autonomy. 13
Therefore, autonomy is measured using two proxy variables – who made the decision for using
contraception, and who decides how to spend money in the household. The autonomy index ranges
from (0-2), with ―0‖ referring to no autonomy while ―2‖ referring to high autonomy regarding
health related decisions.
The reproductive variables that are included in the analysis are mother’s age, birth interval and birth
parity. Maternal age is a continuous variable measured in single years. Birth parity/order is measured
from the first birth to the sixteenth birth. Birth interval is measured in months and is coded into three
dummy categories: ≤24 months, 25-47 months, and ≥48 months. We also control for rural/urban
settings, region of residence, partners’ education, and mothers’ age.
Method of analysis
The dependent variables (complete immunization and child nutrition-height for age) in the analysis
are dichotomous and hence the estimation is done using logistic regression. The logit equation is log
[p / 1- p] = a +bX, where p / (1- p) is the odds of an outcome occurring given the independent
measure of explanatory variable X, and where a indicates the constant and b represents the
coefficient being estimated. Therefore, this equation estimates the log odds of a child being fully
vaccinated and whether the child is stunted. The derived coefficients are interpreted based on their
significance and are then exponentiated to give odds ratios (Bronte-Tinkew & Dejong, 2005), which
show the effect of the independent variables and covariates on the probability of complete
immunization and child stunting.
Given that the 2003 KDHS employed a complex survey design, adjustments were made to account
for clustering, stratification, and unequal weighting. According to An (2002) and Cassell (2006),
accounting for complex survey design in data analysis is important because it 14
allows for estimation of accurate standard errors in cases where the sample has been drawn using
clusters, stratification, and unequal weights.
Findings
Descriptive statistics for child immunization and nutrition
The descriptive statistics presented in Table 1 show that almost half of the children who are between
0-36 months are fully immunized in Kenya. Overall, the level of maternal education is low as well
with only 57% of women having primary education, 23% with secondary education and higher, and
20% of the women having no education at all. Socioeconomic level as measured by the wealth index
variable shows that a higher proportion of women in Kenya are still poor: 44% of women are
categorized as poor; 18% are in the middle rich; and 38% are rich. The results show that the mothers
in this sample have moderate health knowledge as indicated by an average score of 3.4 out of the
total score of seven. The attitude index is low with mothers’ attaining an average score of 2.17 out of
a total score of six. The autonomy score is modest with mothers scoring an average of 0.93 out of a
possible score of two.
The reproductive measures show that 26% of children were born within an interval of two years.
Forty nine of the children have a birth spacing of between 2-4 years, and 25% of children were born
after four years. The average age of mothers was 28 years. Overall, the level of partners’ education is
low, with 48% of men having primary education, 36% with secondary education and higher, and 16%
of the men having no education. Three quarters of mothers live in the rural areas. A majority of the
mothers live in the Rift valley Province (21%) followed by Western (14%). The most common source
of information for mothers was the radio, with 78% reporting they listen to radio. 15
Eighty percent of the children are stunted in Kenya, out of all children who are 0-60 months have
higher proportion of lacking proper nutrition. Overall, among the sample of children (n=5,949)
whose data was used to estimate height for age, their mothers’ level of education can be termed as
low; with 58% of women have primary education, 22% have secondary education and higher, and
20% of the women have no education at all. Socioeconomic level as measured by the wealth index
variable shows that among women with children between 0-60 months, 44% are categorized as poor,
18% are in the middle rich, and 38% are rich. The results show that mothers in this sample have low
levels of health knowledge as indicated by an average score of 2.87 out of the total score of seven.
For the attitude index, the average score was 1.88 out of a total score of six. The autonomy score was
slightly lower with the mothers’ average score being 0.58 out of a possible of two items.
[Insert table 1]
Multivariate Logistic Regression Results (Effect of Maternal Education on Complete Child’s
Immunization)
Table 2 presents the univariate/unadjusted logistic regression (model 1) and multivariate logistic
regression (model 2-8) predicting the effect of maternal education on child health in Kenya. In model
1, children born to mothers with a primary education were 2.17 times more likely to be fully
immunized compared to those who do not have any education at all, p<0.001. In addition, those
children born to mothers with a secondary education were 2.68 times more likely to be fully
immunized compared to those who do not have any education at all, p<0.001. After controlling for
confounding (place of residence, mother’s current age, partner’s education, and region of residence)
i.e. model 2, mother’s education effect attenuated, such that children born to mothers with a primary
education were 1.85 times more likely to be fully immunized 16
compared to those who do not have any education at all, p<0.01. Those children born to mothers with
a secondary education were 2.16 times more likely to be fully immunized compared to those who do
not have any education at all, p<0.01. Furthermore, women from North Eastern, Nyanza, and
Western Provinces were significantly at less odds of having their children fully immunized (OR:0.23,
p<0.001, OR:0.38, p<0.001, and OR:0.59, p<0.05, respectively) compared to those in Nairobi
Province..
Consistent attenuation of the odds ratios was observed through model 5 after controlling for
socioeconomic factor (measured by wealth index), knowledge index, mother’s attitudes index, and
autonomy index. Children born to mothers with a primary education were 1.53 times more likely to
be fully immunized compared to mothers with no education at all, p< 0.05. After further controlling
for reproductive behavior birth interval, and source of information mothers who have completed a
primary education were notably associated with complete immunization of children. Those children
born to mothers with a primary education were 2.20 times more likely to be fully immunized
compared to children born to mothers with no education at all, p<0.05.
[Insert table 2]
Multivariate Logistic Regression Results (Effect of Maternal Education on Child’s Nutritional Status)
Table 3 indicates results predicting the effect of maternal education on child’s nutritional status in
Kenya. Model 1 (unadjusted results), mother’s primary education is significantly related to stunted
growth among children. Children born to mothers with a primary education were at 94% lower odds
of having stunted growth, p<0.01) compared to children born to mothers with 17
no education at all. However, after control for place of residence, mother’s current age, partner’s
education, region of residence, and wealth index (models 2 and 3), there was no significant
association between maternal education and child’s nutritional status. However, after controlling for
knowledge index (model 4) children born to mothers with a primary education were 15% more likely
to have stunted growth compared to those born to mothers with no education at all, p< 0.05. There
was significant association between maternal education and child’s nutritional status after further
controlling for mother’s attitude index, autonomy index, birth order parity, birth intervals, and
information source (model 8)..
[Insert table 3]
Discussion
The objective of this study was to assess the effect of maternal education on child health in Kenya, as
measured by complete immunization and nutritional status. We find that mother’s health knowledge,
receptive attitude toward modern medicine, reading newspapers, and birth interval (25-47 months)
are significantly related to children’s complete immunization. The findings on receptive attitude and
use of modern health care are similar to those of previous studies (Frost et al., 2005; Castro & Juarez,
1995; Glewwe, 1999), which demonstrated that there is an important link between mothers’ attitudes
toward modern medicine, maternal education, and child health. Therefore, we conclude that
transforming the attitudes of mothers in Kenya, is one way of encouraging them to seek
immunization services, hence increasing the number of children who are fully vaccinated against
childhood diseases and illnesses.
Concurrent with previous research (Castro & Juarez, 1995; Glewwe 1999), our findings support the
association between health knowledge and children’s complete immunization. We can conclude that
formal education is important in imparting health knowledge to women, which 18
in turn leads to important improvements in child health. In addition, as indicated in the multivariate
analysis, the significance of reading newspapers as a source of information and maternal education
makes us conclude that education is critical in enhancing women’s understanding and synthesis of
information about health issues. These issues include immunization campaigns appearing in the print
media. We had expected to find a significant relationship between mother’s autonomy and child
health (both complete immunization and stunting) as shown in previous literature. One plausible
reason is the lack of reliable measures of autonomy in the KDHS data set, which could have affected
the results of the current study. In addition, the socioeconomic indicators did not have a significant
impact on children’s immunization status, which is contrary to previous research by Desai and Alva
(1998).
For nutritional status, our findings suggest that the amount of wealth in a household is a primary
driving factor in children’s health. Contrary to our expectations, it is SES and not mother’s education
that predicts children’s nutritional status. We also find surprising results in North Eastern Province,
where the findings suggest that children are less likely to be stunted. Because of the nomadic
lifestyle, harsh living conditions, and the low levels of education associated with this region, we
expected a reverse relationship. We speculate that three factors may be influencing this outcome.
First, ownership of large herds of cattle provides the families with milk, which is a rich source of
nutrients for their children at early ages. Secondly, because of low levels of education, most mothers
in the province do not work outside the home. Thus, the mothers are available to breastfeed their
infants for longer durations. Third, children in North Eastern province may be benefiting from the
food supplies provided by the various government agencies and non-governmental organizations
(NGOs) that operate in the province. Listening to radio was associated with less likelihood of
children being stunted. This means that radio 19
broadcasts are important tools for disseminating public health awareness and information campaigns
in Kenya.
Our results are limited by the absence of direct measures for attitude and autonomy of mothers in
relation to child health; these indices were constructed using proxy and not direct indicators. This
construction might have affected the impact of these variables on the dependent variable. Since this
study was cross–sectional in nature, it was not possible to assess the impact of maternal education on
children’s health over time. In addition, our findings are limited to one context—Kenya; therefore,
cannot be generalized to other African countries. Future research should seek to establish the
determinants of low immunization coverage particularly in Nyanza province, as well as the reasons
why children in North Eastern are not stunted.
Overall, our findings have significant policy implications for child’s health in Kenya. Increasing
levels of health knowledge among women is important in achieving better children health outcomes,
especially complete immunization. One way of improving the knowledge levels is through
incorporating health knowledge into the primary school curricula so as to reach the majority of
children—among them young girls who are future mothers and more likely to be mothers themselves
at a younger age (Magadi, 2006). Transforming mothers’ attitudes toward modern medicine is critical
in ensuring that the majority of children in Kenya are fully immunized against childhood diseases
and illnesses. Thus, targeted information campaigns—through media sources such as the radio and
newspaper that are aimed at changing women’s attitudes toward modern health care should be
implemented in order to educate mothers about the benefits of having their children fully immunized,
as well as the importance of proper nutrition for healthy growth and development. This is a key issue
to reducing the infant/child 20
mortality rates in Kenya, hence, bringing Kenya a step closer to achieving the MDG goal of
improving child health and reducing child mortality.
A closer government attention is required in Nyanza Province to boost the immunization coverage
for children between 12-35 months in the province. These results show that an integrated approach
that includes improving schooling levels, increasing the levels of health knowledge, and developing
positive attitudes toward modern health care among mothers is beneficial for children immunization.
However, improved wealth in the household is a primary driving force behind better nutrition of
children. Thus, targeted policy initiatives aimed at eradicating poverty and malnutrition are critical to
ensuring that children across all the eight provinces have access to nutritious food critical for healthy
growth and development. 21
References
Alam, N. (2000). Teenage motherhood and infant mortality in Bangladesh: Maternal age
dependent effect of parity one. Journal of Biosocial Science, 32, 229-236.
An, A. B. (2002). Performing logistic regression on survey data with the new Surveylogistic
Procedure [cited 2007 May 27th]; Available from:
http://www2.sas.com/proceedings/sugi27/p258- 27.pdf.
Bachrach, C. A., Clogg, C. C., & Carver, K. (1993). Outcomes of early childbearing: Summary
of a conference. Journal of Research on Adolescents, 3, 337-348.
Barrette, H., & Brown, A. (1996). Health, hygiene, and maternal education: Evidence from
Gambia. Social Science & Medicine, 43(11), 1579-1590.
Bastien, J. W. (1992). Drum and stethoscope: Integrating ethno medicine and biomedicine in
Bolivia. Salt Lake City, UT: University of Utah Press.
Becker, G. S. (1964). Human capital. New York: Columbia University Press.
Bicego, G. T., & Ahmad, O. C. (1996). Infant and child mortality. Demographic and health
surveys comparative studies No. 20. Calverton, MD: macro International Inc.
Bicego, G. T. & Boerma J. T. (1993). Maternal Education and child survival: A comparative
analysis of DHS data. (pp 177-204). In Proceedings of the Demographic and Health Survey
World conference, vol.1. Columbia, MD: IRD/Macro International Inc.
Boerma, J. T., Sommerfelt A. E, & Rustein S. O. (1991). Child Morbidity and treatment patterns.
DHS Comparative Studies, No. 4. Columbia, MD:IRD/Macro International, Inc.
Bronte-Tinkew, J., & Dejong, F. G. 2005. Do household structure and household economic
resources predict childhood immunization? Evidence from Jamaica and Trinidad and
Tobago. Population Research and Policy Review, 24, 27-57.
Caldwell, J. C. (1979). Education as a factor in mortality decline: An examination of Nigerian data
Population Studies, 33, 395-413.
Caldwell, J. C., (1986). Routes to low mortality in poor countries. Population and Development
Review, 12(2), 171-220.
Caldwell, J. C., Caldwell, P., Gajanayake, I., Orubuloye, I. O., Pieris, I., & Reddy, D. H. (1990).
Cultural, social and behavioural determinants of health and their mechanisms: A report on related
research programs. In J. C. Caldwell, S. Findley, P. Caldwell, G. Santow, W. Cosford, J. Braid, & D.
Broers-Freeman (Eds), What we know about the health transition: The cultural, social, and
behavioural determinants of health, Vol. 2. (pp. 534-541). Canberra, Australia: The Australian
National University Printing Service.
Caldwell, J. C. (1993). Health transition: the cultural, social, and behavioural determinants of health
in the third world. Social Science and Medicine, 32(2), 125-135.
Caldwell, J. & Caldwell, P. (1993). Roles of women, and communities in preventing illness and
providing health services in developing countries. In J. gribble, & S. Preston (Eds.), The
epidemiological transition: policy planning and implications for developing countries;
workshop proceedings (pp. 252-271). Washington DC. National Academy Press.
Caldwell, J. C. (1994). How is greater maternal education translated into child health? Health
Transition Review, 4, 224-229.
Capital News, Nairobi. Kenya in second polio drive [cited 2009 April 22nd]; Available from:
http://www.capitalfm.co.ke/news/Local/Kenya-in-second-polio-immunisation-drive-
4143.html.
Cassell, D. L. (2006). Wait wait, don’t tell me… you’re using the wrong proc! [cited 2009 22
September 02nd]; Available from: from http://www2.sas.com/proceedings/sugi31/193-
31.pdf.
Casterline, J. B. (Ed.), (2001). Diffusion, processes and fertility transition, Washington, DC:
National Academy Press.
Castro M. T. & Juarez, F. (1995). The impact of women’s education on fertility in Latin
America: Searching for explanations. International Family Planning Perspectives, 21,
52-80.
Central Bureau of Statistics (CBS) [Kenya], Ministry of Health (MOH) [Kenya], and ORC
Macro. 2004. Kenya Demographic and Health Survey 2003. Calverton, Maryland: CBS,
MOH, and ORC Macro.
Cleland, J. (1990). Maternal education and child survival: Further evidence and explanations. In.:
J. Caldwell, S. Findley, P. Caldwell, G. Santow, W. Cosford, J. Braid, & D. Broies-
Freedman (Eds.), What we know about the health transition: The cultural, social, and
behavioural determinants of health, 1, 400-419. Canberra, Australia: The Australian
National University Printing Service.
Cleland, J. Bicego, G. & Fegan, G. (1991). Socio-economic Inequalities in childhood Mortality:
The 1970 compared to the 1980 pp135-54 in Proceedings of the Demographic and Health
Surveys World Conference, Vol 1 Columbia, MD: IRD/ Macro International, Inc.
Cleland, J. & Kauffman, G. (1993). Education, fertility and child survival: Unraveling the links.
Paper Presented at the Workshop of International Union for Scientific Study of the
Population, held in Barcelona, November 10-14.
Cleland, J. G., & Van Ginneken, J. V. (1988). Maternal education and child survival in
developing countries : the search for pathways of influence. Social Science and
Medicine, 27(2), 1357-1368.
Cochrane, S. H. (1980). The effects of education on health. World Bank staff working paper No
405. Washington, DC: World Bank.
Curtis, S. L., & Steele, F. (1996). Variations in familial neonatal mortality risks in four countries.
Journal of Biosocial Science, 28(2), 141-159.
Defo, B. K. (1997). Effects of socio-economic disadvantage and women’s status on women’s
health in Cameroon. Social Science & Medicine, 44(7), 1023-1042.
Desai, S., & Alva, S. (1998). Maternal education and child health: Is there a causal relationship?
Demography, 35(1), 71-81.
Dibley, M. J., Goldsby, J.,Strehling, M., & Trowbridge, F. L. (1987). Development of
normalized curves for the international growth reference: Historical and technical
considerations. American Journal of Clinical Nutrition, 46,736-748.
Frenzen, P. D., & Hogan, D. P. (1982). The impact of class, education and health care on infant
mortality in a developing society: The case of Rural Thailand. Demography, 19(3), 391-
408.
Forste, R. (1994). The effects of breastfeeding and birth spacing on infant mortality in Bolivia.
Population Studies, 48, 497-511.
Frost, M. B., Forste, R., & Haas, D. W. (2005). Maternal education and child nutritional status in
Bolivia: Finding the links. Social Science and Medicine 60, 395-407.
Ghubaju, B. B. (1986). Effect of birth spacing on infant and child mortality in Rural Nepal.
Journal of Biosocial Science, 18(4), 435-447.
Goldman, N., Pebley, A. R., & Gragnolati, M. (2002). Choices about treatment for ARI and
diarrhea in rural Guatemala. Social Science and Medicine, 55, 1693-1712. 23
Glewwe, P. (1999). Why does mother’s schooling raise child health in developing countries?
Evidence from Morocco? The Journal of Human Resources, 34(1), 124-136.
Heaton, T. M., Forste, R. Hoffmann, J. P. & Flake, D. (2005). Cross national variation in family
influences on child health. Social science and medicine 60, 97-108.
Hobcraft, J. N., Mc Donald, J. W. & Rustein, S. O. (1984) ―Socio-economic factors in infant and
child mortality: A cross-national comparison‖ Population Studies, 38,193-223.
Hyman, H. H., C.R. Wright, and J. S. Reed. (1976). The enduring effects of education. Chicago,
IL: University of Chicago Press
Jejeebhoy, S. J. (1995). Women’s education, autonomy, and reproductive behavior: Experience
from developing countries. Oxford: Clarendon Press.
KDHS 2003. Kenya Demographic and Health Surveys, Macro International: Baltimore
Kishor, S. (1995). Gender differentials in child mortality: A review of evidence. In Monica Das
Gupta, Lincoln C. Chen and T. N. Krishnan (eds), Women’s Health in India: Risk and
Vulnerability, Bombay: Oxford University Press
LeVine, R. A., LeVine , S. E., Richman, A., Tapita Uribe, R. M., & Sunderland Correa, C.
(1994). Schooling and survival: The impact of maternal education on health and
reproduction in the third world. In L. C. Chen, A. Kleinman, &N. C. Ware (Eds.), Health
and social change in international perspective (p. 303-338). Boston, MA: Harvard School of Public
Health.
Lindstrom, D. P. (2000). The effects of breastfeeding and birth spacing on infant and early
childhood mortality in Ethiopia. Social Biology, 47(1), 1-17.
Magadi, M. (2006). Poor pregnancy outcomes among adolescents in South Nyanza region of
Kenya. Afr J Reprod Health, 10(1), 26-38.
Manda, S. O. (1999). Birth intervals, breastfeeding and determinants of childhood mortality in
Malawi. Social Science and Medicine, 48, 301-312.
Mason, K. O. (1984). The status of women: A review of its relationship to fertility and mortality.
Paper prepared for the Population Science Division of the Rockefeller Foundation. New
York, NY.
Martin, L. G., Trussel, J., Salvail, F. R., & Shah, N. M. (1983). Covariates of child mortality in
the Philippines, Indonesia, and Pakistan: An analysis based on hazard models. Population Studies,
37(3), 417-432.
Mirowsky, J., & Ross, C. E. (1989). Social causes of psychological distress, New York, NY:
Plenum.
Mirowsky, J. & Ross, C. E. (2003). Education social status and health. Gruyter: New York.
Population Reference Bureau (2008). World Population Data Sheet: Global Demographic Divide
Widens. Washington, D.C.
Ross, C. E., & Mirowsky, J. (1999). Refining the association between education and health: The
effects of quantity, credential, and selectivity. Demography, 36(4), 445-460.
Saraswathi, T. S. (1992). Child survival and health and their linkages with psycho-social factors
in the home and community. Psychology and Developing countries, 4(1), 73-87.
Sewell, W. H. and R. M. Hauser. (1976). Education, occupation and earnings. New York, NY:
Academic Press.
Shawky, S., & Milaat, W. (2001). Cumulative impact of early maternal marital age during the
childbearing period. Pediatric and Perinatal Epidemiology, 15(1), 27-33.
Schultz, T. P. (2002) ―Why Governments Should Invest More to Educate Girls‖ World
Development, 30(2), 207-225. 24
Spaeth, J. L. (1976). Cognitive complexity: A dimension underlying the socioeconomic
achievement process, pp 103-31 in Schooling and Achievement in American Society. Edited by W. H.
Sewell, R. M. Hauser, and D. L Featherman. New York: Academic.
Streatfield, K., Singaribun, M., & Diamond, I. (1990). Maternal education and child
immunization. Demography. 27,(3), 447-455.
Sullivan, J. M., Rustein, S. O. & Bicego, G. T. (1994). Infant and child mortality. Demographic
and health surveys comparative studies no. 15. Calverton, MD: Macro International, Inc
Tagoe-Darko, E. D. (1995). Maternal education and child health and survival in Ghana.
Dissertation, Department of Sociology, Brown University.
United Nations (1985). Socio-economic differentials in child mortality in developing countries.
New York NY: UN Department of International Economic and Social Affairs.
Victoria, C. G., Smith, P. G., Vaughan, J. P. (1986). Social and environmental influences on
child mortality in Brazil: Logistic regression analysis of data from census files. Journal
of Biosocial Science,18(1), 87-101.
Ware, H. (1984). Effects of maternal education, women’s roles, and child care on child mortality.
In W. Mosley, & Chen (Eds.), Child survival: Strategies for research, Population and
development review, a supplement, 10, 191-214.
World Bank. (1993). World development report 1993; Investing in Health, New York: Oxford
University Press.
World Bank (2004). Millennium development goals [cited 2008 January 09th]; Available from:
http://www.ddp-ext.worldbank.org/ext/GMIS.
Young, F.W., Edmonston, B., & Andes, N. (1983). Community level determinants of infant and
child mortality in Peru. Social Indicators Research, 12(1), 65-81.
Zeiltin, M., Ghassemi, H., & Mansour, M. (1990). Positive deviance in child nutrition—with
emphasis on psychosocial and behavioral aspects and implications for development.
Tokyo, Japan: The United Nations University. 25
Table 1: Study summary NUTRITION IMMUNIZATION
descriptive statistic,
KDHS-2003. Study
Characteristics
N=5949 N=2169
Dependent Variables
Child’s nutrition status and Child’s Immunization Status
Stunted (Yes) Immunized 80% 49%
(Yes)
Not Stunted (No) Not 20% 51%
Immunized (No)
Primary Independent Variable
Mother’s education
No education 20% 20%
Primary 58% 57%
Secondary + 22% 23%
Secondary Independent Variables
Socioeconomic status
Household wealth index
Poor 44% 44%
Middle 18% 18%
Rich 38% 38%
Knowledge index
Range 0-7 0-7
Mean 2.84 3.4
(SD) 1.44 1.24
Attitude index
Range 0-6 0-6
Mean 1.88 2.17
(SD) 1.44 1.41
Autonomy
Range 0-2 0-2
Mean 0.58 0.93
(SD) 0.57 0.47
Reproductive Variables
Birth interval
0-24 months 27% 26%
25-27 months 49% 49%
48+ months 24% 25%
Mother’s current age
Mean 28 28
(SD) 6.66 6.5
Birth Parity
Range 1-16 1-16
Mean 3.48 3.53
(SD) 2.46 2.46
Control Variables
Partner’s education level
No education 16% 16%
Primary 49% 48%
Secondary + 35% 36%
Place and type of residence
Rural 74% 75%
Urban 26% 25%
Region of residence
Nairobi 9% 9%
Central 12% 13%
Coast 12% 12%
Eastern 12% 12%
Nyanza 13% 12%
Rift Valley 20% 21%
Western 14% 14%
North Eastern 8% 7%
Sources of Information
Listen to radio
Yes 78% 78%
No 22% 22%
Read newspapers
Yes 32% 33%
No 68% 67%
Watch TV
Yes 28% 27%
No 72% 73%

http://paa2010.princeton.edu/downloa
d.aspx?submissionId=100182
Why does mother's schooling raise child
health in developing countries? Evidence
from Morocco.(includes appendix)
Journal of Human Resources
| January 01, 1999 | Glewwe, Paul | Copyright

Ads by Google

Sport-Related Concussions
Clinical report from The Council on Sports Medicine & Fitness.
www.pediatrics.aappublications.org
I. Introduction

Child health is a key indicator of the quality of life in developing countries. Mother's years of
education is often positively associated with improved child health and nutritional status (see
Behrman, 1990). There are a variety of mechanisms through which mother's education could
raise child health: (1) Direct acquisition of basic health knowledge in school may provide future
mothers with information useful for diagnosing and treating child health problems; (2) Literacy
and numeracy skills learned in school may enhance mothers' abilities to treat child illnesses,
conditional on health knowledge, and also should help mothers increase their stock of health
knowledge after leaving school; and (3) Exposure to modern society in general via schooling
may change women's attitudes toward traditional methods of raising children and treating their
health problems.

This paper attempts to assess the relative importance of these three mechanisms, using the 1990-
91 Moroccan Enquete Nationale des Niveaux de Vie des Menages (ENNVM). Knowledge of the
relative importance of these mechanisms can have important policy implications. For example, if
the main impact of education comes from directly raising mothers' basic health knowledge, such
knowledge should be taught in schools as early as possible (that is, before girls drop out) and
perhaps should also be taught in special education courses for women of child-bearing age who
have already left school.

The paper is organized as follows. Section II reviews, in broad terms, the impact of mother's
education on child health and briefly reviews the recent literature. Section III discusses the data
and the estimation strategy. Section IV presents the empirical results. Section V decomposes the
total impact of mother's schooling on child health. Section VI summarizes the results.

II. Mother's Education and Child Health

A. General Discussion

Figure 1 provides a schematic framework for thinking about the determinants of child health and
nutritional status. As seen at the bottom of that figure, child health is ultimately determined by
three distinct sets of factors: 1. Health and nutritional inputs provided by the household (arrow i);
2. The local health environment (arrow f); and 3. The child's health endowment (arrow h). Health
and nutritional inputs provided by the household include prenatal care, breastmilk, breastmilk
substitutes such as infant formula, calories from adult foods (for weaned children), medicines,
and medical care. The quality of household drinking water sources, toilet facilities, and other
hygienic conditions can also be considered as health and nutritional inputs provided by the
household. The local health environment consists of all community characteristics that directly
affect child health and are generally beyond the control of the parents, such as prevalence of
parasites and the incidence of contagious disease among the general population. Finally, the child
health endowment consists of all components of the child's genetic inheritance that have
implications for his or her health.

Household health and nutritional inputs are determined by household decisions that reflect the
characteristics of the household, the local community, and the child, such as (initial) household
assets, parental schooling, community economic and health-related characteristics (such as the
availability and prices of medical services), and each child's health endowment. This paper
focuses on parental schooling, particularly mother's schooling; father's schooling, apart from its
income effect, is less likely to be important for maintaining children's health.

Schultz (1984) argues that mother's education can influence child health in five ways:(1) (1)
Education may lead to a more efficient mix of health goods used to produce child health; (2)
Better educated mothers may be more effective at producing child health for a given amount and
mix of health goods; (3) Schooling can affect parents' preferences in systematic ways - for
example, educated mothers tend to opt for fewer but healthier children; (4) More schooling
should raise family incomes, either through higher wages or increased productivity in self-
employment, which should improve child health status; and (5) Education raises the opportunity
costs of time, which tends to increase the time mothers spend working outside the home and thus
reduce time for child care - this effect of schooling could reduce child health by reducing both
maternal time devoted to child care and duration of breastfeeding. In Figure 1, the third and
fourth pathways are represented by the arrows a-a[prime] (and also by a-a[double prime]) and
acd (via a-a[triple prime] and c-c[triple prime]), respectively.(2) The first two pathways, which
reflect the direct effect of the health knowledge and cognitive skills that education imparts, have
received little attention in the literature. What is it about schooling that makes mothers more
efficient in producing child health?

Figure 1 presents two mechanisms through which schooling could influence the choice of health
and nutritional inputs via the knowledge and skills it provides.(3) First, schools may directly
teach effective health care practices to students. This pathway is denoted by b-abc. For example,
the impact of diarrhea on child health can be reduced by oral rehydration therapy (ORT), which
can be taught even in primary schools (see Cash 1983). Second, schooling can influence child
health inputs through the cognitive skills imparted, such as literacy and numeracy. Literate
mothers are better able to read written instructions for treating of childhood diseases, and
numeracy enables mothers to better monitor illnesses and apply treatments. This direct effect is
shown by c-c[double prime] in Figure 1. Literacy and numeracy also enable mothers to increase
their health knowledge by enabling them to gather information from written sources. This
indirect effect is path c-c[prime]-abc in Figure 1.

Figure 1 also depicts how factors other than schooling influence child health. Household physical
assets raise household incomes (arrow d), which should have a positive effect on both nutritional
inputs (such as calories) and environmental conditions around the home. The choice of health
and nutritional inputs will also be affected by factors associated with the supply of these inputs in
the community (arrow e). For example, the availability and quality of health and non-health
community facilities affects the decisions households make regarding health and nutritional
inputs. Finally, the child's health endowment will also affect household health and nutritional
inputs (via arrow g), since more sickly children usually receive larger amounts of health and
nutritional inputs.

B. Recent Empirical Evidence

Many recent studies have examined the impact of mother's (and father's) education on child
health. For comprehensive reviews of the literature see Behrman and Deolalikar (1988),
Behrman (1990), and Strauss and Thomas (1995). The discussion here will be limited to an
overview of a few recent studies, focusing on the impact of mother's education on height-for-age
and weight-for-height.

Studies of the determinants of child height and weight in many countries have found positive
effects of mother's education. Most of these studies presented reduced form estimates, but a few
went further, examining the pathways by which mother's education improves child health. In the
Philippines, Barrera (1990) found that better educated mothers tended to wean their children
sooner, but they compensated for this shortened breastfeeding time with better care; overall, their
children were healthier as measured by higher height-for-age z-scores.(4) The only published
study that focuses on the "information processing" attributes of schooling is by Thomas, Strauss,
and Henriques (1991), which used Brazilian data that included variables for whether a woman
reads a newspaper, listens to the radio, or watches television. Mother's schooling was not
significant when dummy variables were included for these "information processing" activities;
the newspaper and radio variables were significant in rural areas but only the television variable
was significant in urban areas.

Among the most interesting studies are those based on the Cebu Longitudinal Health and
Nutrition Survey. Several studies have used these data to model the pathways by which
exogenous variables influence child nutritional status and morbidity. The Cebu Study Team
(1991, 1992) found that mother's education leads to improved waste disposal and higher non-
breastmilk calorie intake, both of which reduce the incidence of diarrhea. Maternal education
also leads to earlier weaning, which can increase episodes of diarrhea, but the net effect of
maternal education is to reduce the incidence of diarrhea.

An important critique of findings that mother's education improves child health is the hypothesis
that education simply reflects unobserved maternal characteristics. Wolfe and Behrman (1987)
used Nicaraguan data on mothers' siblings to control for unobserved family fixed effects. They
found that applying these controls leaves no significant effect of mother's education on child
anthropometric status. However, Strauss (1990) found in Cote d'Ivoire that mother's education
raises child height-for-age and weight-for-height, even after using family fixed effects
estimators.

In summary, there is considerable evidence that mother's education improves child health, and
some evidence on how this occurs. Still, there are no studies that distinguish between the literacy
and numeracy impacts of schooling and other, more general, impacts. Also, there are no studies
that attempt to assess directly the impact of mother's health knowledge on child health.

III. Analytical Framework, Data and Estimation Strategy

A. Analytical Framework

Estimation of the pathways by which mothers' schooling affects child health is not necessarily
straightforward. This subsection provides a framework for thinking about how to estimate these
relationships. Recall Figure 1. The bottom of that figure shows how health and nutritional inputs,
the environment and a child's health endowment jointly determine child health. This can be
expressed in terms of a production function for child health:

(1) [H.sub.i] = f([HI.sub.i], [E.sub.i], [[Epsilon].sub.i])

where [H.sub.i] is the health of child i, [HI.sub.i] is a vector of health inputs chosen by child i's
household, [E.sub.i] is a vector summarizing the environmental conditions surrounding child i,
and [[Epsilon].sub.i] is the child's genetic health endowment. Parents take this technological
relationship into account as best they can when making decisions that affect their children's
health. Although [E.sub.i] and [[Epsilon].sub.i] are outside the household's control,(5) health and
nutritional input choices are chosen by the household.

Estimation of Equation 1 would require detailed information on a large number of health inputs,
which is not feasible with the 1990-91 ENNVM data. However, as seen in Figure 1, one can
substitute out these health inputs and obtain a reduced form relationship that shows how
exogenous variables (those shown at the top of Figure 1) determine child health:(6)

(2) [H.sub.i] = g([FS.sub.i], [MS.sub.i], [HA.sub.i], [E.sub.i], [[Epsilon].sub.i])

where [FS.sub.i] and [MS.sub.i] are father's and mother's schooling, respectively, and [HA.sub.i]
is the initial assets of child i's household.

Although Equation 2 is much easier to estimate, and often has been estimated, it does not
indicate what aspects of mother's schooling lead to improved child health. Referring again to
Figure 1, one can obtain a better understanding of the impact of mother's schooling by replacing
it in Equation 2 with the educational outcomes it directly affects, namely cognitive …

http://www.accessmylibrary.com/articl
e-1G1-54157872/why-does-mother-
schooling.html
The effects of mothers’ education on the nutritional
outcomes of
their children in Nicaragua
Kirk Geale
This manuscript was prepared under the supervision of Professor Louise Grogan,
Department of Economics, College of Management and Economics.
Using data from the 2001 Nicaragua Demographic and Health Survey, this paper examines the relationship of a child’s
nutritional health outcomes relative to the completion of secondary education of their mother by measuring her child’s
height-for-age and weight-for-height. This study focuses on Nicaragua in particular, in contrast to other literature
surveying Latin America as a whole. The persistence of malnutrition amongst the population makes Nicaragua a
candidate for research in this area, especially in face of educational reforms in the country approximately 10 years
prior. In this study the control variables include paternal education, geographic location, socioeconomic status, birth
order, and household size; combined to help attenuate the effects of maternal education. The analysis is subdivided to
examine the relation of mothers’ education to health outcomes for children of each gender. It was found that maternal
secondary education is significant for all scenarios with the exception of gender-separated weight-for-height, and that
there is a stronger correlation between health outcomes for girls than for boys when examining maternal education.
“More education for women would be helpful for our country”21 – Michell Roman, School Teacher in San
Marcos, Nicaragua
he effects of education in relation to outcomes of
children’s health have been well documented in
academic literature. This study looks at the congruency of
these effects on households in Nicaragua. Two statistical
methods are used in this paper to measure child health
outcomes, the first of which is ordinary least-squares
regressions (OLS) and the second is the probit maximum
likelihood model. Both of these models measure height-forage
and weight-for-height, where the independent variable
recording maternal secondary education is of primary
interest. The probit model measures the likelihood of an
event occurring (here, the event is stunting or wasting)
whereas OLS uses independent variables to explain the
health outcomes of the children in the study and how those
outcomes change as the explanatory variables change.
A study published in 2004 with regard to this same topic
in Ethiopia found that parental education is one of the “key
determinants of chronic child malnutrition in Ethiopia.”1 This
study also found that the effect of maternal education is
almost double that of paternal education and that mothers’
secondary school education has a significant effect on
anthropometric scores of their children when compared to
uneducated mothers.
Uthman’s 2009 study examined the effects of
socioeconomic status on child health in Nigeria. While
Uthman found that monetary wealth alone was the most
prevalent contributing factor, accounting for one third of the
inequality in malnutrition, with mothers’ education playing a
lesser role,23 this study predicts that the effects of maternal
education will be stronger than wealth. This is due in part to
the high rates of return on education reported in Latin
America. These two studies use many of the same variables
to measure nutritional inequality, with results that are similar
in direction but different in magnitude.
Nicaragua is a candidate for current research due to the
persistence and prevalence of high levels of malnutrition
despite educational reforms that have occurred in its recent
history. The Frente Sandinista de Liberacion’s (Sandinista)
governmental regime from 1979 to 1990 saw a radical
change in educational policy. It began with the famous
literacy campaign in 1980, significantly reducing the levels
of illiteracy. A longitudinal study conducted on the effects of
this campaign indicates that women who became literate as
adults now have healthier children.14 Following the literacy
initiative came an institutionalized education program in
which many of the participants would be the appropriate age
for inclusion in the paternal and maternal demographics in
this paper. Although many of the overarching goals of the
program were not achieved, the educational revolution boasts
improvements such as significantly increased levels of
enrollment in all educational categories (primary, secondary,
tertiary) and significantly increased literacy rates. In spite of
these successes, malnutrition endures in Nicaragua; Arnove
and Dewees concluded this was in part due to the lack of

Adult education and child nutrition: the role

of family and community
1. H Moestue,
2. S Huttly

+ Author Affiliations

1. Nutrition and Public Health Intervention Research Unit, London School of Hygiene and
Tropical Medicine, London, UK
1. H Moestue, London School of Hygiene and Tropical Medicine, London, UK;
helenmoestue@gmail.com

 Accepted 21 March 2007

Next Section

Abstract
Background: It is well established that mothers’ education has positive effects on child nutrition
in developing countries. Less explored is the effect exerted by the education of other individuals
—mothers’ friends, neighbours and family.

Objectives: To examine independent effects of mothers’, fathers’ and grandmothers’ education

on child height-for-age and weight-for-age z-score, and the role of community-level maternal
literacy over and above parental education and other individual-level factors.

Methods: Cross-sectional data were analysed for 5692 children from Andhra Pradesh State in
India and Vietnam sampled within “sites” (20 from each country) and then within “communities”
(31 from Vietnam and 102 from India). Multilevel regression analysis was undertaken to account
for confounders and geographical clustering of observations.

Results: Child nutrition is positively and independently associated with mothers’, fathers’ and
grandmothers’ education. The association with grandmothers’ education was statistically
significant in the India sample only and was stronger for boys: the adjusted mean difference in
height-for-age z-scores between boys living with an educated grandmother and those not was
0.64 (95% CI 0.29 to 0.99, p<0.001). In the Vietnam sample, child nutrition was associated with
the proportion of literate mothers in the community, adjusting for parental education and other
confounders (height 0.81, 95% CI 0.29 to 1.31, p = 0.002).

Conclusion: The results imply that an individual-level perspective may fail to capture the entire
impact of education on child nutrition, and support a call for a widening of focus of nutrition
policy and programmes from the mother–child pair towards the broader context of their family
and community.

Previous SectionNext Section

It is well established that mothers’ education has a positive effect on child nutrition in developing
countries. In school, girls can acquire skills that are later used to access modern health services
and comprehend health messages. Less explored, however, is the effect exerted by the education
of other individuals—mothers’ friends, neighbours and family—who may influence child
nutrition directly (though childcare) or indirectly (though modification of the effect of maternal
education).

Despite recent improvements, approximately half of preschool children remain underweight in

Asia, which is the highest level in the world; in India 62 million children are underweight and in
Vietnam there are 5 million.1 2 In both countries there has been concern that economic advances
have not been reflected in improvements in child nutrition.3–5 Education is viewed as a key
element in the overall strategy for reducing malnutrition in the developing world. This view is
based on evidence spanning four decades for a positive effect of adult education on child health
and nutrition in both developed and developing countries.6–10 The effect is particularly strong
for female education. In India, education has been shown to explain the spatial clustering of
malnutrition and death,11 12 and even low levels of education have been shown to increase child
survival prospects and health-related behaviours.13

The positive effect of maternal education is, nevertheless, not universal.14 In Vietnam, a recent
study using two nationally representative surveys found no significant effect of maternal
education on child nutrition.4 5 This is perhaps because increased pressure on women to work
outside the home has had negative consequences for childcare and breastfeeding practices, which
may have outweighed any potential benefit of education.

It has been argued that children’s greater “exposure time” to mothers explains why the effect of
education is greater for mothers than fathers, as fathers tend to work outside the home.15 If,
however, employment draws mothers away from childcare at home, the presence of other
household members who are better educated than the mother may offer the potential for an
improvement in the overall quality of childcare. This argument is supported by other studies
drawing attention to the role played by older siblings and grandparents in childcare.12–16

The concept of “mass education” was first introduced by Caldwell6 in 1980 and developed by
Cleland and Jejeebhoy,17 who used survey data from south Asia to demonstrate that in
communities where the average level of education is high, the fertility of the women with little or
no education is lower than would otherwise be expected. Kravdal18–20 claimed that an
expansion of education would reduce mortality, not only because more women would enter an
educational category associated with lower mortality, but also because all community members,
including those who themselves remained uneducated, would benefit from the generally higher
level of education in the community. These views are also shared by others.21–23

A review of the literature reveals that most studies of adult education and child nutrition have
focused on parents only. Few have examined the effect of education among other household or
community members, although recent research suggests that these “others” may play an
important role in determining child nutrition.20–23 The objectives of this study are therefore to
explore the effect of mothers’, fathers’ and grandmothers’ education on child nutrition, and to
estimate, using multilevel methods, the effect of community-level maternal literacy on child
nutrition over and above parental education and other individual-level factors.

Previous SectionNext Section

METHODS
Study design

This study uses cross-sectional data from the first round of the Young Lives study (in 2002) in
Vietnam and Andhra Pradesh State in south India. Full details of the Young Lives study are
available at www.younglives.org.uk. The data from the first round of the Young Lives study are
available from the UK Public Archive, study no. SN5307 (www.esds.ac.uk). In each country
approximately 3000 children were sampled from 20 sites: 2000 children aged 6 to 17.9 months at
enrolment (“one-year olds”) and 1000 children aged 7.5 to 8.5 years (“eight-year olds”). Sentinel
sites, and the communities within the sites, were selected semipurposively. Households were
selected using a method equivalent to random sampling. Detailed information on the study
design and sampling methods has been published elsewhere.24–26 Although the data are not
nationally representative, we will hereafter refer to the data being from “India” and “Vietnam”.

Main variables

Anthropometric indices were used as indicators of child nutrition, following procedures

recommended by the World Health Organisation.27–29 Children were classified as “stunted” (an
indicator of chronic malnutrition) if they have height-for-age z-scores below −2 SD and
“underweight” (indicator of both chronic and acute malnutrition) if they have weight-for-age z-
scores below −2 SD.28 29 Education was measured by asking the respondent what level of
formal schooling each household member had completed. The data were recoded into the four
main categories: “none”, “primary” (corresponding to ages 6–11 years), “secondary” (11–14/15
years) or “higher” (15 years or above). For grandmothers, all education categories were
combined because only a small number of grandmothers had secondary or higher education.
Community-level maternal literacy was measured as the proportion of mothers who were literate
in the community, as estimated from the sample (data on paternal literacy was not available).
Literacy was assessed by asking “Can you read and understand a letter or newspaper easily, with
difficulty, or not at all, in any language?”. Answers were coded “easily”, “with difficulty” or “not
at all”. Women who answered “easily” were categorised as literate. The choice of a community-
level literacy variable, rather than an education variable, is based on the hypothesis that literacy
is the key outcome of education and that the “externality” of education is spread mainly through
literacy mechanisms.30

Statistical methods

For descriptive analysis we used chi-squared tests, Student’s t-tests and F-tests to assess the
statistical significance of differences between proportions, two means or more than two means,
respectively. Multivariable regression analysis was used to adjust simultaneously for multiple
confounders. Confounders were selected on the basis of UNICEF’s model for the determinants of
malnutrition, which incorporates biological and socioeconomic causes at both micro and macro
levels.31 Statistical interactions were assessed by entering a dummy interaction term into the
model. Exact p values are presented in the text and statistical significance was assumed at the 5%
level, except for statistical interactions, which used the 10% level. A test for trend was
undertaken by applying the likelihood ratio test between a model in which the categorical
variable is specified as continuous and a model excluding this variable altogether. A three-level
model is used to account for the hierarchical structure of the data,32 33 in which level 1 refers to
measurements of the individual child and the child’s household, level 2 the community and level
3 the site. Data were analysed for 5692 children/households in which the mother was the main
caregiver of the child; this included 1946 one-year olds from Vietnam and 1899 from India, 963
eight-year olds from Vietnam and 884 from India, and excluding cases with missing values for
any of the variables selected for use in the analysis (N  =  327).

Ethical clearance was obtained from participating research institutions in the United Kingdom
and each Young Lives study country. Before interview, informed consent was obtained from
participants.

Previous SectionNext Section

RESULTS
The prevalence of stunting and underweight was 19% and 24% in Vietnam and 28% and 43% in
India. The between-country differences correspond to what has been shown elsewhere, with
Vietnamese children being better nourished on average than Indian children.34 Table 1 shows the
distribution of maternal, paternal and grandmother education in each country. It is apparent that
adults have a higher level of schooling in Vietnam than in Andhra Pradesh (p<0.001), which also
corresponds to previous literature.35

View this table:

 In this window
 In a new window

Table 1 The level of schooling completed by mothers, fathers and grandmothers in India and Vietnam
Objective 1: maternal education

Tables 2 and 3 show, for height and weight-for-age, respectively, the crude effect of maternal
education on child nutrition (model A), and the effect adjusting for confounders (model B) as
well as independent risk factors (model C). The tables show that the crude effect diminishes
somewhat after adjustment for potential confounders and risk factors, although it remains
statistically significant and linear for both nutrition outcomes (p<0.001). There was no difference
in the education effect by country or child sex, but it was stronger in urban areas than rural
(height p = 0.079, weight p<0.001) and stronger among the relatively wealthy households than the
poor (height p = 0.094, weight p = 0.006). Differences in the effect of education were, however,
found between one and eight-year olds in terms of weight-for-age (p<0.001), but not height-for-
age (p = 0.849), with the effect on weight being greater for the one-year olds.

View this table:

 In this window
 In a new window

Table 2 Regression output: the crude and adjusted association between maternal education and child
height-for-age z-scores (N  =  5692)
View this table:

 In this window
 In a new window

Table 3 Regression output: the crude and adjusted association between maternal education and child
weight-for-age z-scores (N  =  5692)

Objective 2: paternal and grandmother education

Table 4 shows that fathers’ education is positively associated with nutrition, and that whereas the
presence of a grandmother in the household is not associated with child nutrition, her education
has a positive effect on height-for-age (p = 0.053) and weight-for-age (p<0.001), adjusting for
parental education and other confounders. Further analysis was undertaken to explore the role of
grandmothers’ education, as this is a relatively new area of interest within the literature. First, it
was found that the effect was stronger in urban areas than rural (height p = 0.028, weight p = 
0.027), as observed earlier for maternal education. Second, it was found that the effect was only
evident in India and was stronger for boys than girls. In India the mean difference between boys
having an educated grandmother in the household and boys who did not was 0.64 height-for-age
z-scores (95% confidence interval (CI) 0.29 to 0.99, p<0.001) and 0.52 weight-for-age z-scores
(95% CI 0.25 to 0.78, p<0.001). The difference was smaller for Indian girls (0.07, 95% CI −0.24
to 0.36, p = 0.671 for height-for-age, and 0.29, 95% CI 0.07 to 0.52, p = 0.010 for weight-for-age)
and was statistically non-significant for both sexes in Vietnam.

View this table:

 In this window
  In a new window

Table 4 Regression output: the effect of grandmothers’ presence and education on child height and
weight-for-age z-scores in all households (N  =  5464)

Objective 3: community-level education

Tables 5 and 6 show, for height and weight-for-age, respectively, the crude results (model A), the
results adjusted for potential confounders and risk factors but excluding parental education
(model B), and the results adjusted for all confounders and risk factors, including parental
education (model C). Adjustment for non-education confounders and independent risk factors
diminishes the effect estimates for community-level maternal literacy, although the effect
remains statistically significant for both height-for-age (p = 0.002) and weight-for-age (p = 0.031).
Once parental education has also been accounted for, however, the positive effect is reduced; the
statistical significance of the effect disappears for child weight-for-age (p = 0.496) but remains
borderline for height-for-age (p = 0.073). Once stratified by country, the results reveal that the
effect is only significant in Vietnam (Vietnam 0.81, 95% CI 0.29 to 1.31, p = 0.002; India 0.12,
95% CI −0.31 to 0.55, p = 0.577). The lack of an effect in India may be explained by the
particular definition of a “community”, which in India was “a small village or a cluster of
hamlets” and the small sample size of households within some of the communities sampled.

View this table:

 In this window
 In a new window

Table 5 Regression output: the crude and adjusted association between community-level maternal
literacy and height-for-age z-scores (N  =  5474)
View this table:

 In this window
 In a new window

Table 6 Regression output: the crude and adjusted association between community-level maternal
literacy and weight-for-age z-scores (N  =  5474)
Previous SectionNext Section

DISCUSSION
The study has demonstrated a positive association between child nutrition—both stunting and
underweight—and the education of adults living within the child’s household and community.
Particularly interesting, given the lack of previous evidence, is the positive association shown
between child nutrition and grandmothers’ education in India and community-level maternal
literacy in Vietnam.
What this paper adds

 It is well established that mothers’ education has a positive effect on child nutrition in
developing countries. Less explored, however, is the effect exerted by the education of
other individuals—mothers’ friends, neighbours and family—who may influence child
nutrition directly (though childcare) or indirectly (though modification of the effect of
maternal education).

 This study shows that children’s nutrition is associated with the education of many adult
household and community members. Particularly interesting, given the lack of previous
evidence, is the finding that child nutrition in one sample (India) is positively associated
with grandmothers’ education, and that in the other sample (Vietnam), child nutrition is
positively associated with community-level maternal literacy. The findings imply that an
individual-level perspective may fail to capture the entire impact of education on child
nutrition, and support a call for a widening of focus of nutrition policy and programmes
from the mother–child pair towards the broader context of their family and community.

Policy implications

There are two major policy implications from this research. First it supports existing efforts to
universalise basic education and adult literacy programmes, as education is positively associated
with child nutrition. The second implication is more novel: policies currently aimed at improving
child nutrition may benefit from recognising key actors other than the mother. The common
approach of targeting women of reproductive age for health and nutrition interventions may
overlook “influential others” in the family or community. Now that a growing number of women
in India and Vietnam are taking up employment outside the household, the role of other carers
such as grandmothers and siblings is likely to become increasingly important in determining
child nutrition. Therefore the findings support a call to widen the focus of nutrition policy and
programmes from the mother–child pair towards the broader context of their family and
community.

The positive association with grandmothers’ education in India is plausible given the
overwhelming evidence of health benefits of female education, and given what we know about
grandmothers’ responsibilities in childcare.36 It supports previous studies that have highlighted
the role of grandmothers in determining health and fertility behaviour, whether “positively” or
“negatively”, and which call for a greater participation of grandmothers in behavioural-change
interventions.36–38

The positive association between child nutrition and community-level maternal literacy in
Vietnam, over and above parental education and other individual-level confounders and risk
factors, suggests that children may benefit from living in literate communities. Again, this is
highly plausible. Despite the economic transition of the past 20 years, Vietnam’s social structure
is still largely influenced by the socialist planned economy. The literacy of other members of a
“commune” can therefore realistically affect childcare through the sharing of knowledge and
imitation of behaviour. Educated people often create demand for services and infrastructure and
tend to have more political clout than less educated populations.23
This research adds to the overwhelming evidence for a positive effect of adult education on child
nutrition. Education-effects were similar for both height and weight-for-age z-scores, suggesting
they operate over the medium to longer term. The study also shows—in contrast to some other
studies in the past—that fathers’ education is associated with child nutrition, and that this
association was as strong as mothers’ education.38 It is plausible that fathers, often heading the
households, make decisions and behave in ways that directly or indirectly affect childcare.

The difference in findings between India and Vietnam may be explained by differences in the
cultural significance placed on the extended family or commune/community. For example,
grandmothers in India and communes in Vietnam are known to exert strong influences on
childcare, so it is not unexpected that this study finds an effect of their education. Between-
country differences may also be the result of methodological factors, such as the size and
structure of the two datasets. For example, the absence in India (in contrast with Vietnam) of an
effect of community-level maternal literacy may be explained by the small size of the
communities selected—this argument is supported by findings from a recent nationwide study in
India showing a clear positive effect of community-level education on child mortality.20
Similarly, the lack of an effect of grandmothers’ education in Vietnam (in contrast with India)
may be explained, partly at least, by the small number of grandmothers included in the sample.
Both methodological and cultural factors may thus lie behind the between-country differences
observed. Furthermore, one should bear in mind, when interpreting the results, that the data were
cross-sectional, meaning that causality cannot be proved and that further longitudinal research
would be necessary to corroborate the findings.

Despite these weaknesses, this small study has succeeded in linking a number of key issues of
interest among educationalists, economists and nutritionists, and highlighting their synergies.
Within the disciplines of education and economics, the issue of education “externalities” is
currently much debated, with increasingly sophisticated methods being developed to capture the
social “knock-on” effects from investments in education.30 40 The findings presented here
demonstrate that externalities of education are observed both within and between households,
that the externalities can be measured through child anthropometry, and that an individual-level
perspective may fail to capture the entire impact of education on child nutrition.

There are two implications for nutrition policy emerging from this research. First it supports
existing efforts to universalise basic education and adult literacy programmes, as education is
positively associated with child nutrition. The second implication is more novel: policies
currently aimed at improving child nutrition may benefit from recognising and including key
actors other than the mother. It is plausible that the common approach of targeting women of
reproductive age for health or nutrition interventions may overlook “influential others” in
childcare and therefore miss useful avenues of influence. This study shows that mothers play a
role in determining child nutrition, as well as fathers, grandmothers and the community at large.
This is a timely finding as there is an increasing realisation within the field of nutrition that
changing mothers’ behaviour through targeted health education is difficult without
simultaneously tackling underlying societal issues and intra-household power relations, which
are known to influence mothers’ decision-making and actions.37 41 42 Moreover, as Asian
countries undergo globalisation, and increasing numbers of mothers take up employment outside
the home, other carers such as grandmothers and siblings are likely to play an increasingly
important role in determining child nutrition and health. We therefore support a call to widen the
focus of nutrition policy and programmes from the mother–child pair towards the broader
context of their family and community, including mothers’ social networks as well as locally
influential individuals such as teachers and community leaders.43

Further research is needed to corroborate the findings and interpretations presented here, in
particular to improve our understanding of the mechanisms that underlie the relationship between
child nutrition and adult education, including education measured at the community level. In
addition, programmatic research is needed to explore practical ways in which to involve
influential household and community members including mother-in-laws, teachers and local
leaders, in nutritional interventions to increase programme effectiveness.

Previous SectionNext Section

Acknowledgments
Access to the Young Lives data via the UK data archive is gratefully acknowledged.

Previous SectionNext Section

Footnotes
 Funding: The research studentship of HM was funded by the Medical Research Council.
 Competing interests:None.

Previous Section

REFERENCES
1. ↵
1. de Onis M,

2. Frongillo EA,

3. Blossner M

. Is malnutrition declining? An analysis of changes in levels of child malnutrition since 1980. Bull
WHO 2000;78:1222–33.

[Medline][Web of Science]

2. ↵
1. Mason J,
 2. Hunt J,

3. Parker D,

4. et al

. Investing in child nutrition in Asia. Asian Dev Rev 1999;17:1–32.

3. ↵
1. Claeson M,

2. Bos ER,

3. Mawji T,

4. et al

. Reducing child mortality in India in the new millennium. Bull WHO 2000;78:1192–9.

[Medline][Web of Science]

4. ↵
1. Thang NM,

2. Popkin B

. Child malnutrition in Vietnam and its transition in an era of economic growth. J Hum Nutr Diet
2003;16:233–44.

[CrossRef][Medline]

5. ↵
1. Glewwe P,

2. Koch S,

3. Nguyen BL

. Child nutrition, economic growth and the provision of health care services in Vietnam in the
1990s Policy research working paper series 2776. Washington, DC:World Bank 2002.

6. ↵
1. Caldwell J

. Mass education as a determinant of the timing of fertility decline. Popul Dev Rev 1980;6:225–
55.

[CrossRef][Web of Science]
7.
1. Cleland JG,

2. Van Ginneken JK

. Maternal education and child survival in developing countries: the search for pathways of
influence. Soc Sci Med 1988;27:1357–68.

[CrossRef][Medline][Web of Science]

8.
1. Cochrane SH,

2. O’Hara D,

3. Leslie J

. The effects of education on health Washington, DC:World Bank 1980.

9.
1. Frost MB,

2. Forste R,

3. Haas DW

. Maternal education and child nutritional status in Bolivia: finding the links. Soc Sci Med
2005;60:395–407.

[CrossRef][Medline]

10. ↵
1. Bicego GT,

2. Boerma JT

. Maternal education and child survival: a comparative study of survey data from 17 countries.
Soc Sci Med 1993;36:1207–27.

[CrossRef][Medline][Web of Science]

11. ↵
1. Griffiths P,

2. Matthews Z,

3. Hinde A
 . Gender, family, and the nutritional status of children in three culturally contrasting states of
India. Soc Sci Med 2002;55:775–90.

[CrossRef][Medline]

12. ↵
1. Das Gupta M

. Death clustering, mothers’ education and the determinants of child mortality in rural Punjab,
India. Popul Stud (Camb) 1990;44:489–505.

13. ↵
1. Basu AM,

2. Stephenson R

. Low levels of maternal education and the proximate determinants of childhood mortality: a
little learning is not a dangerous thing. Soc Sci Med 2005;60:2011–23.

[CrossRef][Medline][Web of Science]

14. ↵
1. White H,

2. Masset E

. Child poverty in Vietnam: using adult equivalence scales to estimate income-poverty in

Vietnam. Young Lives 2002. www.younglives.org.uk (accessed 1 June 2005).

15. ↵
1. Bishai D

. Quality time: how parents’ schooling affects child health through its interaction with childcare
time in Bangladesh. Health Econ 1996;5:383–407.

[CrossRef][Medline][Web of Science]

16. ↵
1. Van Esterik P

. Care, caregiving, and caregivers. Food Nutr Bull 1995;16:378–88.

17. ↵
1. Jeffrey R,

2. AM Basu
 3. Cleland J,

4. Jejeebhoy SJ

. Maternal schooling and fertility: evidence from census and surveys. In: Jeffrey R, AM Basu, eds.
Girls’ schooling, women’s autonomy and fertility change in South Asia New Dehli: Sage
Publications, 1996.

18. ↵
1. Kravdal O

. A search for aggregate-level effects of education on fertility, using data from Zimbabwe.
http://www.demographic_research.org/Volumes/Vol3/3 (accessed 1 June 2005).

19.
1. Kravdal O

. Education and fertility in sub-Saharan Africa: individual and community effects. Demography
2002;39:233–50.

[Medline][Web of Science]

20. ↵
1. Kravdal O

. Child mortality in India: the community-level effect of education. Popul Stud (Camb)
2004;58:177–92.

[Medline]

21. ↵
1. Moursund A,

2. Kravdal O

. Individual and community effects of women’s education and autonomy on contraceptive use in
India. Popul Stud (Camb) 2003;57:285–301.

22.
1. McNay K,

2. Arokiasamy P,

3. Cassen R
 . Why are uneducated women in India using contraception? A multilevel analysis. Popul Stud
(Camb) 2003;57:21–40.

[Medline]

23. ↵
1. Alderman H,

2. Hentschel J,

3. Sabates R

. With the help of one’s neighbors: externalities in the production of nutrition in Peru. Soc Sci
Med 2003;56:2019–31.

[CrossRef][Medline]

24. ↵
1. Wilson I,

2. Huttly SR

. Young Lives: a case study of sample design for longitudinal research. Working paper no. 10.
Young Lives 2004. www.younglives.org.uk (accessed 1 June 2005).

25.
1. Galab S,

2. Gopinath MR,

3. Antony P,

4. et al

. Young Lives preliminary country report: Andhra Pradesh India. Young Lives 2003.
www.younglives.org.uk (accessed 1 June 2005).

26. ↵
1. Tuan T,

2. Lan PT,

3. Harpham T,

4. et al

. Young Lives preliminary country report: Vietnam. Young Lives 2004. www.younglives.org.uk
(accessed 1 June 2005).
27. ↵
1. Hamill PV,

2. Drizd TA,

3. Johnson CL,

4. et al

. Physical growth: National Center for Health Statistics percentiles. Am J Clin Nutr 1979;32:607–
29.

[Abstract/FREE Full text]

28. ↵

World Health Organization. Use and interpretation of anthropometric indicators of nutritional

status. Bull WHO 1986;64:929–41.

[Medline][Web of Science]

29. ↵

World Health Organization. Physical status: the use and interpretation of anthropometry
Geneva:World Health Organization 1995.

30. ↵
1. Basu K,

2. Narayan A,

3. Ravallion M

. Is literacy shared within households? Theory and evidence for Bangladesh. Labour Econ
2001;8:649–65.

[CrossRef]

31. ↵

UNICEF. The state of the world’s children New York:UNICEF 1998.

32. ↵
1. Goldstein H

. Multilevel modelling of survey data. The Statistician 1991;40:235–44.

 [CrossRef]

33. ↵
1. Rabash J,

2. Browne W,

3. Goldstein H,

4. et al

. A user’s guide to MLwiN London:Institute of Education 2000.

34. ↵
1. de Onis M,

2. Monteiro C,

3. Akre J,

4. et al

. The worldwide magnitude of protein-energy malnutrition: an overview from the WHO Global
Database on Child Growth. Bull WHO 1993;71:703–12.

[Medline][Web of Science]

35. ↵

UNESCO. Comparing education statistics across the world. Montreal, Canada: UNESCO Institute
for Statistics, Global Education Digest, 2004.

36. ↵
1. Sharma M,

2. Kanani S

. Grandmothers’ influence on child care. Indian J Pediatr 2006;73:295–8.

[Medline]

37. ↵
1. Aubel J,

2. Toure I,

3. Diagne M
 . Senegalese grandmothers promote improved maternal and child nutrition practices: the
guardians of tradition are not adverse to change. Soc Sci Med 2004;59:945–59.

[CrossRef][Medline]

38. ↵
1. Hobcraf JN,

2. McDonald JW,

3. Rutstein SO

. Socio-economic factors in infant and child mortality: a cross-national comparison. Popul Stud
(Camb) 1984;38:193–223.

[Medline]

39.
1. Sear R,

2. Mace R,

3. McGregor IA

. Maternal grandmothers improve nutritional status and survival of children in rural Gambia.
Proc Biol Sci 2000;267:1641–7.

[Abstract/FREE Full text]

40. ↵
1. Gibson J

. Literacy and intrahousehold externalities. World Dev 2001;29:155–66.

[CrossRef]

41. ↵
1. Save the Children UK

. Thin on the ground: questioning the evidence behind World Bank-funded community nutrition
projects in Bangladesh, Ethiopia and Uganda London:Save the Children 2003.

42. ↵
1. Good B
 . Medicine, rationality, and experience: an anthropological perspective Cambridge; New
York:Cambridge University Press 1994.

43. ↵
1. Moestue H,

2. Huttly SR,

3. Sarella L,

4. et al

. ‘The bigger the better’ – mothers’ social networks and child nutrition in Andhra Pradesh. Public
Health Nutr 2007;10:1274–82

[Medline]

J Epidemiol Community Health 2008;62:153-159 doi:10.1136/jech.2006.05857

Influence of Maternal Education on Child

Immunization and Stunting in Kenya
B. A. Abuya, E. O. Onsomu, J. K. Kimani and D. Moore

 Download PDF (198.9 KB)

 View HTML
 Permissions & Reprints

 Related
 Journal

View Related Documents

 no access

Journal Article

Inequities in Under-Five Mortality in Nigeria: Differentials by Religious Affiliation of the

Mother

Diddy Antai, Gebrenegus Ghilagaber, Sara Wedrén, Gloria Macassa and Tahereh Moradi
 Journal of Religion and Health, 2009, Volume 48, Number 3, Pages 290-304

o Show SummaryHide Summary

o Download PDF (188.9 KB)

o View HTML

Journal ArticleInequities in Under-Five Mortality in Nigeria: Differentials by Religious

Affiliation of the MotherDiddy Antai

 full access

Journal Article

Online First™Open Access

Child mortality in rural India

Bas van der Klaauw and Limin Wang

Journal of Population Economics, Online First™, 2 December 2009

o Show SummaryHide Summary

o Download PDF (458.0 KB)

o View HTML

Journal ArticleChild mortality in rural IndiaBas van der Klaauw

 no access

Journal Article

Maternal Education and Child Healthcare in Bangladesh

Mohammed Nazmul Huq and Tarana Tasnim

Maternal and Child Health Journal, 2008, Volume 12, Number 1, Pages 43-51

o Show SummaryHide Summary

o Download PDF (261.0 KB)

o View HTML

Journal ArticleMaternal Education and Child Healthcare in BangladeshMohammed

Nazmul Huq
 no access

Journal Article

Online First™

Maternal Health Literacy and Late Initiation of Immunizations Among an Inner-City

Birth Cohort

Susmita Pati, Kristen A. Feemster, Zeinab Mohamad, Alex Fiks and Robert Grundmeier,
et al.

Maternal and Child Health Journal, Online First™, 24 February 2010

o Show SummaryHide Summary

o Download PDF (227.9 KB)

o View HTML

Journal ArticleMaternal Health Literacy and Late Initiation of Immunizations Among an

Inner-City Birth CohortSusmita Pati

 no access

Journal Article

Impact of Prenatal Care on Infant Survival in Bangladesh

Rathavuth Hong and Martin Ruiz-Beltran

Maternal and Child Health Journal, 2007, Volume 11, Number 2, Pages 199-206

o Show SummaryHide Summary

o Download PDF (212.6 KB)

o View HTML

Journal ArticleImpact of Prenatal Care on Infant Survival in BangladeshRathavuth Hong

 no access

Journal Article

Measuring socioeconomic status in health research in developing countries: Should we be

focusing on households, communities or both?

Jean-Christophe Fotso and Barthelemy Kuate-Defo

 Social Indicators Research, 2005, Volume 72, Number 2, Pages 189-237

o Show SummaryHide Summary

o Download PDF (368.7 KB)

Journal ArticleMeasuring socioeconomic status in health research in developing

countries: Should we be focusing on households, communities or both?Jean-Christophe
Fotso

 no access

Journal Article

Determinants of infant and child mortality in Kenya: an analysis controlling for frailty
effects

D. Walter Rasugu Omariba, Roderic Beaujot and Fernando Rajulton

Population Research and Policy Review, 2007, Volume 26, Number 3, Pages 299-321

o Show SummaryHide Summary

o Download PDF (236.6 KB)

o View HTML

Journal ArticleDeterminants of infant and child mortality in Kenya: an analysis

controlling for frailty effectsD. Walter Rasugu Omariba

 no access

Book Chapter

Infant Mortality

Parul Christian

Nutrition and Health, 2008, Nutrition and Health in Developing Countries, Pages 87-111

o Show SummaryHide Summary

o Download PDF (463.9 KB)

o View HTML

o Look Inside

Book ChapterInfant MortalityParul Christian

  no access

Journal Article

Which factors explain the decline in infant and child mortality in Matlab, Bangladesh?

Lauren Hale, Julie DaVanzo, Abdur Razzaque and Mizanur Rahman

Journal of Population Research, 2009, Volume 26, Number 1, Pages 3-20

o Show SummaryHide Summary

o Download PDF (245.6 KB)

o View HTML

Journal ArticleWhich factors explain the decline in infant and child mortality in Matlab,
Bangladesh?Lauren Hale

 no access

Journal Article

Determinants of Birthweight; Gender Based Analysis

Samia Halileh, Niveen Abu-Rmeileh, Graham Watt, Nick Spencer and Nahida Gordon

Maternal and Child Health Journal, 2008, Volume 12, Number 5, Pages 606-612

o Show SummaryHide Summary

o Download PDF (249.1 KB)

o View HTML

Journal ArticleDeterminants of Birthweight; Gender Based AnalysisSamia Halileh

Scroll upScroll down

  References (34)
 Export Citation
 About

Abstract
In 2003, the child mortality rate in Kenya was 115/1000 children compared to 88/1000 average
for Sub-Saharan African countries. This study sought to determine the effect of maternal
education on immunization (n = 2,169) and nutritional status (n = 5,949) on child’s health.
Cross-sectional data, Kenya Demographic Health Survey (KDHS)-2003 were used for data
analyses. 80% of children were stunted and 49% were immunized. After controlling for
confounding, overall, children born to mothers with only a primary education were 2.17 times
more likely to be fully immunized compared to those whose mothers lacked any formal
education, P < 0.001. For nutrition, unadjusted results, children born to mothers with primary
education were at 94% lower odds of having stunted growth compared to mothers with no
primary education, P < 0.01. Policy implications for child health in Kenya should focus on
increasing health knowledge among women for better child health outcomes.

Keywords Child health - Maternal education - Immunization - Child nutrition - Health

knowledge - Kenya

Fulltext Preview
http://www.springerlink.com/content/j201m676n1025682/

The effects of mothers' education on nutritional outcomes of their children in Nicaragua

Kirk Geale

This manuscript was prepared under the supervision of Professor Louise Grogan, Department of
Economics, College of Management and Economics.

Abstract
Using data from the 2001 Nicaragua Demographic and Health Survey, this paper examines the
relationship of a child’s nutritional health outcomes relative to the completion of secondary
education of their mother by measuring her child’s height-for-age and weight-for-height. This
study focuses on Nicaragua in particular, in contrast to other literature surveying Latin America
as a whole. The persistence of malnutrition amongst the population makes Nicaragua a candidate
for research in this area, especially in face of educational reforms in the country approximately
10 years prior. In this study the control variables include paternal education, geographic location,
socioeconomic status, birth order, and household size; combined to help attenuate the effects of
maternal education. The analysis is subdivided to examine the relation of mothers’ education to
health outcomes for children of each gender. It was found that maternal secondary education is
significant for all scenarios with the exception of gender-separated weight-for-height, and that
there is a stronger correlation between health outcomes for girls than for boys when examining
maternal education.

“More education for women would be helpful for our country” [21]. – Michell Roman, School
Teacher in San Marcos, Nicaragua

Introduction

The effects of education in relation to outcomes of children’s health have been well documented
in academic literature. This study looks at the congruency of these effects on households in
Nicaragua. Two statistical methods are used in this paper to measure child health outcomes, the
first of which is ordinary least-squares regressions (OLS) and the second is the probit maximum
likelihood model. Both of these models measure height-for-age and weight-for-height, where the
independent variable recording maternal secondary education is of primary interest. The probit
model measures the likelihood of an event occurring (here, the event is stunting or wasting)
whereas OLS uses independent variables to explain the health outcomes of the children in the
study and how those outcomes change as the explanatory variables change.

A study published in 2004 with regard to this same topic in Ethiopia found that parental
education is one of the “key determinants of chronic child malnutrition in Ethiopia" [1]. This
study also found that the effect of maternal education is almost double that of paternal education
and that mothers’ secondary school education has a significant effect on anthropometric scores of
their children when compared to uneducated mothers.

Uthman’s 2009 study examined the effects of socioeconomic status on child health in Nigeria.
While Uthman found that monetary wealth alone was the most prevalent contributing factor,
accounting for one third of the inequality in malnutrition, with mothers’ education playing a
lesser role [23], this study predicts that the effects of maternal education will be stronger than
wealth. This is due in part to the high rates of return on education reported in Latin America.
These two studies use many of the same variables to measure nutritional inequality, with results
that are similar in direction but different in magnitude.

Nicaragua is a candidate for current research due to the persistence and prevalence of high levels
of malnutrition despite educational reforms that have occurred in its recent history. The Frente
Sandinista de Liberación’s (Sandinista) governmental regime from 1979 to 1990 saw a radical
change in educational policy. It began with the famous literacy campaign in 1980, significantly
reducing the levels of illiteracy. A longitudinal study conducted on the effects of this campaign
indicates that women who became literate as adults now have healthier children [14]. Following
the literacy initiative came an institutionalized education program in which many of the
participants would be the appropriate age for inclusion in the paternal and maternal
demographics in this paper. Although many of the overarching goals of the program were not
achieved, the educational revolution boasts improvements such as significantly increased levels
of enrollment in all educational categories (primary, secondary, tertiary) and significantly
increased literacy rates. In spite of these successes, malnutrition endures in Nicaragua; Arnove
and Dewees concluded this was in part due to the lack of time, stability, and resources in the
country [3]. This paper will show that although this reform was not as successful as the
participants hoped it would be, it was not a misguided vision because the status of Nicaraguan
children’s health is in fact linked to the levels of education of their mothers.

Despite the extensive literature relating maternal education to nutritional outcomes in developing
countries in general, there is still some controversy surrounding such findings. For example, Alva
and Desai endeavoured to show whether or not there is in fact a strong causal relationship
between maternal education and child health due to skepticism on the part of some academics.
The study suggests that it is not appropriate to make a sweeping statement that mothers’
education has an independently strong and positive impact on the survival of their children.
However, the study also notes that in some areas of Latin America the returns to secondary
education (social and private benefits) are noteworthy, and will be the focus of this study in the
country of Nicaragua [2]. Alva and Desai also cite a lack of proper control variables for
community effects on the education coefficients, as well as controlling for income, which will
also be corrected for in the models in this paper.

There are many studies that make generalized predictions about Latin America regarding
maternal education and child health, which are all valid in their own right. This study will
contribute to the existing literature by looking solely at evidence from Nicaragua. Using
econometric techniques, this paper will explore whether maternal education is strongly related to
a child’s health in Nicaragua. This paper does not aim to show causality; only that a mother’s
education level is linked to the well being of her children. In this paper, maternal education is
measured using the completion of secondary education to show whether or not she is educated,
and child health is measured using height-for-age and weight-for-height comparisons. The levels
of education Nicaraguan parents have on average, measured by secondary education, are quite
low. Out of the 6836 observations of mothers, only 24.58% (1680 observations) have completed
secondary education. Of the fathers of children measured, 20.25% (1384 observations) have been
educated [10]. The analysis consists of these variables as well as a variety of reasonable control
variables that will be further explained. As a developing country in Central America, Nicaragua
is a country with much potential. Health status is strongly linked to developing this potential,
both for individuals and for the country’s economic growth as a whole, and therefore it is
important to understand which variables comprise the factors influencing a child’s health in
Nicaragua.

Methods
The data used in this study include data from the 2001 Nicaragua Demographic and Health
survey. This data set is a comprehensive questionnaire with many controls to ensure accurate
representation of the population. The data in use in this study include birth order, gender, age,
household size, proxies of the wealth index, an urban/rural indicator, and provincial dummy
variables. Anthropometric indices including height-for-age and weight-for-age were also
computed and used as the dependent variables. Anthropometry is an effective way to measure
“changes in body dimensions [that] reflect[s] the overall health and welfare of individuals and
populations" [8]. The two dependent variables being measured in this study, namely weight-for-
height (wasting) and height-for-age (stunting) together provide an indicator of overall health
status. This is because wasting measures acute under-nutrition, and is useful “when exact ages
are difficult to determine” and “is appropriate for examining short-term effects" [8]. Stunting on
the other hand, “is an indicator of past growth failure” and manifests from “a slowing in the
growth of the fetus and the child" [8].

The controls used in this experiment were selected because of their influence on child health
outcomes and their identification in existing literature. Household size was included as a control
because it suggests that the household has a certain amount of disposable income available and
therefore will distribute it in limited proportions. It follows that for a fixed amount of income, a
larger portion of income can be spent on the nutrition of each child, the fewer people there are in
the household. This assumes of course, that the children in the household are not working and do
not generate income that can benefit the household.

Birth order is included as a control variable because the amount of attention that the parents are
able to pay to the wellbeing of each child must be divided between all of the existing children.
The more siblings a child has, the less time the parents have to monitor and engage in the child’s
nutrition levels. Also due to the phenomenon of “unwanted children,” the parents may invest
more in children higher in the birth order with the expectation that these resource allocations
may not be divided between other unexpected children they are not yet aware of. Birth control is
not a well understood and utilized method of family planning and so birth order may have a
negative impact on the well being of children with a lower birth order. Parents are not necessarily
actively discriminating amongst their children; they may be unable to reallocate resources over
the long run to offset the advantages of children with higher birth order [16].

The wealth index is comprised of a group of variables that record whether a family owns a toilet,
telephone, bicycle, car, truck, electricity, and a floor in the house. As a whole, this is a necessary
control because the higher the household’s wealth, more income the family is able to spend on
health related purchases, whether it be nutritional food or doctor access. These factors clearly
affect the parents’ ability to properly feed their children. The dummy variables for the wealth
index proxies are constructed in a way that records a 1 if the variable is “true” for the family.
This can be understood in two ways: 1 if the family has a car, 0 if they do not. In another
example, the variable is recorded as 1 if the family has “nofloor” (it is true that they have no
floor), and 0 if they do have a floor.

Urban areas in Nicaragua, as in many other countries, differ greatly in their access to a variety of
products and services, one of which is food and the quality of that food. Because infrastructure in
developing countries is usually underdeveloped, there is a shortage of sustenance required for
proper nutrition and information on proper nutritional practices that families can partake in.
During a visit to a rural area outside of San Marcos for example, I learned that women generally
believe that sugar is needed to keep the fluid around the brain healthy. As a result, children from
a very young age are fed excessive amounts of sugar. The educational infrastructure in rural
areas has been well documented to be inferior of that in urban centers [17].

Dummy variables for each of the country’s 19 provinces were included in order to take into
account differing geographic advantages or disadvantages. As an example, areas that farm
particular crops that do not add to nutrition of the local citizens will not have the benefits
realized by an area that grows nutritious crops. The regional variables help control for the verity
that “poorer children are more likely to live in regions with disadvantageous characteristics"
[23]. They also help to control for the community effects that can be present, for example
spillover in a province with a university, or the presence of a health focused NGO. There are
many community effects, and the provincial dummy variable will help to control for these as
well as the geographic effects. For example, “educated mothers tend to live in more
economically developed areas” and areas that are rich enough to have schools are also probably
rich enough to have medical facilities [2].

When examining coefficients in this paper, it must be kept in mind that the dependent variables
measuring the z-scores [a] for height-for age and weight-for-height have been limited to z-scores
with standard deviations below 5, to eliminate extreme values in the data sets, which are likely
incorrect.

The two anthropometric measures, height-for-age and weight-for-height, were created from the
data in the 2001 Nicaragua Demographic and Health Survey in conjunction with the 2000 United
States CDC growth charts. The CDC Growth Chart is one of two standard measures, and was
chosen over the 1990 UK British growth charts due to geographic proximity, and because the
surveys are only two years apart (CDC 2000 Growth Charts and 2001 Nicaragua Health and
Demographic Survey). Because gender is included within the z-scores for the height-for-age and
weight-for-height variables, it does not appear explicitly as a control variable in any of the
specifications. In the creation of the z-scores for the height-for-age and weight-for-height
anthropometric measures, there are a variety of variables included. One of these is gender; the
others are weight, height, age, and the standards from the CDC growth charts.

The regressions presented in this paper show the independent variables affecting height-for-age
and weight-for-height. There are 19 proxy provincial variables present, which will be represented
in the regression simply as “PROV,” with the coefficient β7j where j is each individual province
(1 £ j £ 19). This method has also been applied to the proxies of the wealth index, which will
appear together as “WEALTH” where the coefficient is β6i where i is each wealth index proxy (1
£ j £ 7). These proxies will only be broken out when there is a specific reason for identifying an
individual component.

The measures of height-for-age and weight-for-height are z-scores, are measured as such in the
first two OLS regressions. In the probit regressions, the z-scores are represented by binary
variables in which 1 represents a z-score from -5 to -2 (stunted or wasted), and 0 represents a z-
score from -2 to 5 (healthy). This conversion was applied to both the height-for-age and weight-
for-height variables to be used in the probit model.

Analysis

Ordinary Least Squares (OLS) Regression

The following equation shows the general form of OLS regressions:

[1] Y = β0 + β1X1i + β2X2i + … + βKXKi + e

Where Y is the dependent variable, X is an independent variable, and β is the magnitude of the
effect of X on Y. The following specification uses this model to measure the z-scores of height-
for-age (Z_HFA) with the aforementioned independent variables.

[2] Z_HFA = β1(F_EDU) + β2(M_EDU) + β3(BIRTH_ORDER) + β4(HHSIZE) +Β5(URBAN)

+ β6i(WEALTH i) + β7j(PROV i) + e

Table 1 illustrates the coefficients on each independent variable.

The results of the regression show that mothers’ education (F_EDU) is statistically significant
[b]. This was expected in the regression analysis, as was the significant effect of fathers’
education (M_EDU) on child health. The coefficients on the maternal and paternal variables are
quite different however; maternal education has more than twice the effect on child health
outcomes when compared to paternal education.

Within the provincial dummy variables, there was a considerable range of statistical significance
as well as in the magnitude of coefficients. The coefficients ranged from -0.24 to 0.79. Generally,
the stronger coefficients pertained to variables that were statistically different from 0. Both
household size and birth order were significant, affecting the measured z-score at a rate of
-0.0416 and -0.0085 respectively.

In the second regression measuring weight-for-height (Z_WFH), the results were quite different.
The specification for this measurement includes the same variables measured against weight-for-
height:

[3] Z_WFH = β1(F_EDU) + β2(M_EDU) + β3(BIRTH_ORDER) + β4(HHSIZE) +

Β5(URBAN) + β6i(WEALTH) + β7j(PROV) + e

From Table 2 it is easy to see the statistical insignificance of almost all of the regressed variables.
The only one with statistical significance is the variable of most importance to this study,
maternal education.

A second set of regressions was run on both height-for-age and weight-for-height after
disaggregating male and female children. All of the independent variables were the same, in
order to determine whether maternal education has a different effect on health outcomes of her
children based on their gender. The results in Table 3 illustrate the differences between the four
combinations of gender and anthropometric measure.
The effect of maternal secondary education on both male and female children found earlier in the
Table 1 regression against height-for-age was reported as .3501295 (1% CI) and in the Table 2
regression against the z-score for weight-for-height maternal education was reported as .1031464
(5% CI). However, due to the overlapping standard error of maternal education variable between
gender-separated OLS specifications, a Chow test was conducted (Appendix 5 [see
supplementary file]) with the null hypothesis that the coefficients are simultaneously equal to
zero. P > F was reported as zero for both the height-for-age and weight-for-height regressions,
indicating that the coefficients are comparable across specifications. It can then be concluded
that daughters’ nutritional outcomes are more strongly related to mothers’ secondary educational
attainment than sons’ in terms of height-for-age, and sons’ outcomes are slightly more responsive
to maternal secondary educational attainment than daughters’ outcomes in terms of weight-for-
height.

It was anticipated that there would be collinearity [e] between certain variables, particularly
mothers’ education and fathers’ education. At 36%, the correlation was lower than expected. Due
to the importance of considering paternal education in a theoretically sound specification, it was
decided to maintain the paternal education variable for inclusion in the specifications. Variance
inflation factors (VIF’s) were measured using 5 as an acceptable maximum to check for
multicollinearity, and none of the VIF’s were greater than 5. In fact, the highest VIF is 2.90, and
the mean was 1.72. It should also be noted that the highest 11 VIF’s were provincial dummy
variables.

The very low R squared [f] measure of goodness of fit points toward the possibility of omitted
variables, and therefore impure heteroskedasticity [g]. To test this hypothesis, the Breusch-
Pagan/Cook-Weisberg test was employed on the regression in Table 4.a against the z-score for
height-for-age, and it was found that there was in fact a high probability of heteroskedasticity.

From the results in Table 4.a it was concluded that robust regressions needed to be used. Also, by
using Cameron and Trivedi’s decomposition of assumptions on the regression errors, the
information matrix in Table 4.b confirms that there are violations on these standard assumptions
of least square regressions. To rectify this, the regression was performed using robust analysis,
resulting in the t statistics recorded in Tables 1, 2, and 3. Adding robustness to the analysis did
not significantly change the t-statistics of any of the independent variables. In the regression
against the z-score for weight-for-height, similar tests for heteroskedasticity were performed,
with the following results.

Tables 5.a and 5.b together reveal conflicting conclusions: from the Cook-Weisburg test, the null
hypothesis of homoskedasticity would not be rejected, yet the Cameron and Trivedi
decomposition would suggest the regression suffers from heteroskedasticity. In order to err on
the side of caution, robust regressions were also used for the weight-for-height regression. With
similar results to the height-for-age regression, the robust analysis changed the t-score of any
given variable by very little, or resulted in no change at all. No variables’ significance at the 95
percent confidence interval was affected. Checking for multicollinearity by comparing VIF’s in
this regression is the same as for the height-for-age regression, because the independent control
variables are the same in both analyses.
Binomial Probit Model
The maximum-likelihood probit model is used to determine the likelihood of a child being
stunted or wasted using the same variables as in the OLS the regressions. The model takes the
form:

[4]

Where i is the probability that the child is 2 standard deviations or below, s is a standard
normal variable, and Zi is the sum of the individual components of the regression.

The resulting coefficient of mothers’ education is -0.3386, with statistical significance. The
interpretation of this negative coefficient is related to the construction of the binary independent
variable: If the child is stunted the independent variable is 1, if not, 0. Therefore mothers’
education significantly reduces the probability of a child being stunted. Paternal education is also
negatively correlated to the probability of stunting, and the wealth index behaves in a similar
manner. Higher birth order also contributes to the likelihood of stunting, as does an urban
location. As expected in this regression, the constant term is statistically significant with a
coefficient of -0.5233. The reason for this is the same as in the least-squares regression; the
lower health status in Nicaragua in comparison to the United States.

The probit analysis measuring the probability of weight-for-height exhibits the same low level of
statistical significance as the least-squares regression. The statistical significance of maternal
education is low, and clearly outside the 95% significance level. This implies that the variable
does not affect the probability of being 2 standard deviations below the mean U.S. weight-for-
height. In fact, the only significant variables are 2 wealth index proxies and 2 provincial dummy
variables. The constant term is significant, with a coefficient of -1.6860, which is consistent with
the reasons previously mentioned. The results of the two probit specifications are displayed in
Appendices 3 and 4 [see supplementary file].

Discussion

It is interesting that in both the probit and OLS regression using weight-for-height the urban
location of a household contributes to lower values of the anthropometric measures. A possible
explanation for this outcome is because a large portion of rural inhabitants farm, and therefore
have access to a primary source of food.

Comparing the two least-squares regressions of height-for-age and weight-for-height produces

some interesting results. Maternal education is significant for both measures, and is in fact the
only variable that spans the two regressions in terms of significance. As mentioned in the
introduction, height-for-age is a longer term measure of child health than weight-for-height,
which has a tendency to be a short-term measure. Further, phenomenon emphasizes the value of
mothers’ education, by showing its importance in both of these circumstances. Maternal
education in this model is the only predictor of both short-term and long-term child health, to the
exclusion of all other independent variables.
Maternal education exhibits clear correlation with child health as was hypothesized, consistent
with generalized findings in past research. This trend is exhibited in Figure 1, showing the
correlation of mothers’ secondary education and z-scores for children’s height-for-age with a
95% confidence interval.

It is easy to see from Figure 1 the positive relationship between maternal education and height-
for-age in Nicaragua. The regression demonstrated the statistical significance of these findings.
In the weight-for-height correlation the values of maternal education fit the z-scores as shown in
Figure 2. A comparison of these graphs visually depicts the larger variability of the maternal
education variable: P>|t| = 0.000 for height-for-age whereas P>|t| = 0.043 for weight-for-height.
This simply means that maternal education predicts child health outcomes measured by stunting
more accurately than for wasting because the regression for stunting has a smaller confidence
interval. It was also expected that the strength of the maternal education variable would be higher
if there were more detailed information regarding the quality of education, as opposed to the
measurement used here which is simply enrollment. Educational quality is difficult to quantify,
but it is logical to assume that with a higher quality of education, the magnitude of maternal
education would be higher.

The constant terms in the two robust regression analyses were expected to be negative. This is
because, as a developing country, it is expected that Nicaragua has a lower average nutritional
attainment than the standard against which it is measured, in this case U.S. children. The constant
term is near what was expected for the height-for-age specification, with a value of -1.4. This
implies that the average child in this sample population is 1.4 deviations below that of
comparable children in the sample U.S. population (in which the mean is 0). However, the
constant term in the weight-for-height regression is reported at 0.03. The statistical insignificance
of this intercept term (P=0.754 > |t|) suggests that variation from this trend is likely and may still
be less than 0. Although there is insufficient evidence to make this conclusion, it is suspected that
it is true and the constant term is in fact less than 0. The reason for the discrepancy between
constant terms in both the least-square regression and probit models of weight-for-height and
height-for-age is simply that the mean of the z-score for the height-for-age regression is -1.2957
and the mean of the z-score for the weight-for-height regression is -0.0662. Higher means for
height-for-age are present in all four models presented in this paper.

Difference in ethnic demographics in Nicaragua likely plays a role in controlling for the effects
influencing the anthropometric measures of Nicaraguan children. The largest demographic is
69% mezito, indicating room for variety in ethnicity [24]. This study has not explicitly controlled
for race or ethnicity, which are thought of as “important…individual-level determinants of child
health" [2]. However, the provincial variables are expected to adequately control for this, as race
varies with geography [7].

An influence upon health statuses in Nicaragua that has not been accounted for in the models
presented in this paper is political affiliation. Inhabitants of a given area are inclined towards a
particular political party and when that particular party is in power, political support is an
advantage due to the distribution of government food. This is true in other arenas as well;
supporters of the political group in power gain other general advantages that influence health.
However, from direct observation it is common for entire geographic areas to be inclined
towards the same party, so it is likely that the geographic variables picked up a large portion of
this effect. If a statistic recording political affiliation were available with the other survey
information, in theory it would help explain the overall fit of the model and improve the
accuracy of measurement in these specifications.

The very low value of the R squared goodness of fit in the least squares regressions (12.35% for
height-for-age and 2.53% for weight-for-age) was surprising. It is likely caused in some part by
specification errors, possibly due to the racial and ethnic variables or political affiliation that
have already been discussed. In the context of this model however, the way R squared is
measured will inevitably be lower than desired due to the nature of the variables. For example, if
a family does not have a truck, there is still a possibility that the mother has secondary education.
Although they both contribute to higher anthropometric z-scores, they are not mutually
exclusive. This principle likely affects a variety of variables. That being said, however, the
almost nonexistent value of the R squared suggests there are one or more significant factors
missing in the regressions.

The pseudo R squared reported in the probit models are not of great concern, except in the
context of the evaluation the two models against each other. The modeling software STATA uses
McFadden’s pseudo R squared measure that “without context has little meaning" [22]. In
comparing the two models however, the height-for-age probit is a better fit than the weight-for-
height model, as was the trend with the OLS regressions.

The single proxy for the wealth index that spans both OLS regressions in terms of 95%
significance is the variable for the truck ownership. This is not surprising as it is the highest
socioeconomic indicator in the wealth proxies. This simply suggests that wealth is correlated
with the health outcomes of the children in the family, and supports previous research that led to
the inclusion of these control variables in the specifications in this paper. The dummy variables
for provinces 7, 10, 11, and 19 were also significant at 95% in the two main OLS regressions,
suggesting that these regions influence child health outcomes. However, the discrepancies
between positive and negative coefficients influencing the two health measures portray
ambiguous results.

Conclusions

Maternal education has a significant effect on the health outcomes of a mother’s male and female
children when measured by both height-for-age and weight-for-height. However, the relation of
maternal secondary education on female children’s height-for-age is larger than male height-for-
age. The probit model reveals that maternal secondary education is significantly negatively
correlated to the probability of a child being stunted, but not significantly related to the wasting
of her child. This study reveals that in Nicaragua, a mother’s completion of secondary education
correlates positively with higher levels of child health.

Acknowledgments

The motivation for this research was driven by a volunteer trip to a rural region of Nicaragua
called San Marcos. During my trip I lived, worked, and got to know the people in the community
near the Skylark Center, all of whom contributed to a very inspirational month long journey.
Thanks are due to Karl and Heather Murch, Michell Roman, Joni, and Elvin for their help in the
creation of this paper – and of course to my supervisor, Professor Louise Grogan.

Endnotes

a) A z-score is the number of standard deviations a set of data is away from the mean (average).
The standard deviation shows how much variation there is between an observation and the mean.
b) Significant within the confidence interval expressed in the Table, that is, there is an x %
chance the variable’s coefficient will be outside the value of β ± the corresponding standard
error.
c) The other variables explaining the model are the 19 provincial variables. The full specification
can be seen in Appendix 1 [see supplementary file].
d) The additional explanatory variables are the 19 provincial variables, as well as the wealth
index proxies (the same as are listed in Table 1). The full specification can be seen in Appendix 2
[see supplementary file].
e) A violation of the OLS assumptions where two explanatory variables are correlated.
f) R squared is a measurement of the overall fit of the specification, that is the variability of the
dependent variable that can be accounted for by the independent variables.
g) A violation of the OLS assumptions where the variance of the error term is not consistent.

References

1. Alderman, Harold; Luc, Christiaensen. (2004). Child Malnutrition in Ethiopia: Can

Maternal Knowledge Augment the Role of Income? Economic Development and
Cultural Change, 52, 287-312.
2. Alva, Soumya; Desai, Sonalde. (1998). Maternal Education and Child Health: Is There a
Strong Causal Relationship? Demography, 35(1), 71-81.
3. Arnove, Robert; Dewees, Anthony. (1991). Education and Revolutionary Transformation
in Nicaragua, 1979-1990. Comparative Education Review, 35(1), 92-109.
4. Bannon, Ian; Marques, José. (2003). Central America: Education Reform in a Post-
Conflict Setting, Opportunities ad Challenges. CPR Working Papers: Social
Development Department, Paper No. 4, 1-25.
5. CARMEN Country Profiles. (2006). WHO Global InfoBase Online. Retrieved from
https://apps.who.int/infobase/report.aspx?rid=111&iso=NIC&goButton=Go.
6. Cataldo, M., Filipp, L., Finney, J., Kidwell, S., Mellits, D., Quaskey, S., … Starfield, B.
(1993). Determinants of Children’s Health Care Use: An Investigation of Psychological
Factors. Medical Care, 31(9), 767-783.
7. Charles, Camille Z. (2003). The Dynamics of Racial Residential Segregation. Annual
Review of Sociology, 29, 167-207.
8. Cogill, Bruce. (2001). Anthropometric Indicators Measurement Guide. Food and
Nutritional Assistance Project. (Academy for Educational Development). Washington,
DC, 8-96.
9. Conger, Rand; Lorenz, Frederick; Elder, Glen; Wickrama, K. (1998). Parental Education
and Adolescent Self-Reported Physical Health. Journal of Marriage and Family, 60(4),
967-978.
10. Demographic and Health Surveys. Nicaragua: Standard DHS, 2001 [Stata System Files].
Retrieved from http://www.measuredhs.com/accesssurveys/.
11. Esquivel, Yannete; Gurdian, Nidia Veronica; Handa, Sudhanshu; Lopez, Blancadilia;
Pineda, Heiling; Regalia, Ferdinando. (2009). Non-formal basic education as a
development priority: Evidence from Nicaragua. Economics of Education Review, 28,
512-522.
12. Fuller, Bruce; Rivarola, Magdalena. (1998). Nicaragua’s Experiment to Decentralize
Schools: Views of Parents, Teachers, and Directors. Impact Evaluation of Education
Reforms: Working Paper No. 5, 1-87.
13. Glewwe, Paul. (1998). Why does Mother’s Schooling Raise Child Health in Developing
Countries? Evidence from Morocco. The Journal of Human Resources, 34(1), 124-159.
14. Hanemann, Ulrike. (2006). Nicaragua’s Literacy Campaign. Paper commissioned for the
EFA Global Monitoring Report 2006, Literacy for Life, 1-13.
15. Henriques, Maria-Helena; Thomas, Duncan; Strauss, John. (1991). How Does Mother’s
Education Affect Child Height? The Journal of Human Resources, 26(2) 183-211.
16. Horton, Susan. (1988). Birth Order and Child Nutritional Status: Evidence from the
Philippines. Economic Development and Social Change, 36(2), 341-354.
17. Kahn, Mahmood. (2001). Rural Poverty in Developing Countries: Implications for Public
Policy. International Monetary Fund: Economic Issues, 26. Retrieved from
http://www.imf.org/external/pubs/ft/issues/issues26/index.htm.
18. Murch, Karl, personal communications, May, 2009.
19. Pefia, Rodolfo; Persson, Lars-Ake. (2000). The Effect of Poverty, Social Inequity, and
Maternal Education on Infant Mortality in Nicaragua, 1988-1993. American Journal of
Public Health, 90(1), 64-69.
20. Riding, Alan. (1980, June 3). Nicaragua Drafts the Young for a War on Illiteracy. New
York Times, p. A2.
21. Roman, Michell, personal communications, May, 2009.
22. UCLA: Academic Technology Services, Statistical Consulting Group. (2006). FAQ: What
are pseudo R-squareds? Retrieved from
http://www.ats.ucla.edu/stat/mult_pkg/faq/general/Psuedo_RSquareds.htm.
 23. Uthman, Olalekan. (2009). Decomposing Socio-economic inequality in childhood
malnutrition in Nigeria. Maternal and Child Nutrition, 5, 358-367.
24. The World Factbook. (2009). Central America and Caribbean: Nicaragua. Central
Intelligence Agency. Retrieved from https://www.cia.gov/library/publications/the-world-
factbook/geos/nu.html.

Figures
Tables
http://www.criticalimprov.com/index.php/surg/article/viewArticle/1094/1665

he role of maternal literacy and nutrition knowledge in determining children's

nutritional status

Parul Christian, Rita Abbi, Sunder Gujral, and Tara Gopaldas

Introduction

A mother is the principal provider of the primary care that her child needs during the first six
years of its life. The type of care she provides depends to a large extent on her knowledge and
understanding of some aspects of basic nutrition and health care. It is understandable that her
educational status has been reported to influence her child-care practices.

During the past decade, evidence has accumulated from several studies that maternal education is
an important determinant of infant and child mortality [1, 2]. Chen [2] proposed that children
born of educated mothers have a lower mortality risk because educated women tend to marry and
have their first child at a later age than uneducated women. They also are likely to be more
assertive and to play a greater part in intra-family decision making in favour of their children's
needs. Their husbands tend to be economically better off than those of uneducated women.
Educated mothers may also make earlier and more effective use of health services. It may be
postulated that mothers' education would affect their children's nutritional status by similar
mechanisms, and various studies have shown some degree of association between mothers'
education and the nutritional status of children [3-5].

It cannot be assumed, however, either that the mothers of malnourished children are necessarily
ignorant or that all illiterate mothers. whether their children are healthy or malnourished, are
ignorant [6]. Their knowledge of child nutrition and child-care practices can be expected to have
a significant bearing on their children's nutritional status, but conflicting results have been
reported in this regard. Whereas some studies have observed a positive relationship between
childhood malnutrition and maternal knowledge and beliefs regarding nutrition [7, 8]. others
have shown no such relationship [9, 10].

In the present study we investigated the effects of mothers' literacy status and nutrition
knowledge on the nutritional status of children. We attempted to determine whether literate
mothers had better nutrition knowledge and to elicit specifically the impact of mothers' nutrition
knowledge on their children's nutritional status, controlling for their literacy status and for family
income, which is a well-established factor affecting child nutrition status [8, 11].

Materials and methods

The study was carried out in rural and tribal villages of Panchmahals district of Gujarat state,
India, as part of a baseline survey for the USAID-assisted Integrated Child Development
Services project.

A pretested questionnaire was used to collect information on the education level and nutrition
knowledge of 2,723 mothers. Complete information was not available for all the women, and
therefore only 2,665 were included in the analysis.

Educationally, the women were categorized simply as literate or illiterate. as only 6% had any
formal education and only 3.2% had studied beyond the primary level.
The women's nutrition knowledge was evaluated on a scale of 0 to 9 on the basis of their
responses to a set of nine questions, with one point given for each correct answer and a
maximum possible score of 9. The knowledge components tested were the proper age to
introduce solid foods into a child's diet, types of solid foods to introduce, frequency of feeding,
diet during diarrhoea, management of diarrhoea, awareness of oral rehydration therapy,
preparation of oral rehydration solution, causes and treatment of protein-energy malnutrition,
interpretation of growth charts, and perception of own child's nutritional status. As a majority of
the women had a score of 0 or a very poor score, a score of 4 or above was considered good. The
scores were divided into three categories: 0, 1-3, and 4 or above.

The per capita monthly income of the households of the study population was determined by
interviewing the head of the family or any other senior member in his absence. The families were
categorized as above or below the poverty line, using Rs 65 (US$5) per capita per month as the
dividing point, as suggested by the Gujarat Directorate of Health [12].

All the children of mothers in the study 0-72 months old for whom complete weight, height, and
age data were obtained [13] - 4,004 out of a total 4,242 children - were included in the analysis.
The children were weighed to the nearest 0.1 kg on a Salter-type spring balance. Height was
measured to the nearest 0.1 cm using a portable measuring board; supine length was measured
for children 0- 12 months old and standing height for those one year old or over. The ages of the
children were determined by thoroughly questioning the mothers with the aid of a local events
calendar to assist recall.

The cut-off points used were those suggested by the Indian Academy of Pediatrics [14] for
weight for age (>80%, normal; 71-80%, first-degree malnutrition; 61-70%, second-degree
malnutrition; 51-60%, third-degree malnutrition; <51%, fourth-degree malnutrition), and by
Waterlow [15] for weight for height >=90%, normal; 80-89%, mild malnutrition; 70-79%,
moderate malnutrition; <70%. severe malnutrition) and for height for age (>=95%, normal; 90-
94%, mild malnutrition; 85-89%, moderate malnutrition; <85%, severe malnutrition).

Chi-square values were calculated to determine whether the nutritional status of the children
depended on the mothers' (i) literacy and (ii) nutrition-knowledge score, and to study the
association between these two independent variables and their individual relationship to the
income status of the family. Analysis of covariance was done to study the effect of the mothers'
nutrition knowledge on the children's nutritional status, controlling for their literacy status. The
per capita monthly income was also included in the model to determine whether the association
between maternal nutrition-knowledge scores and child nutritional status was still significant
after testing for the confounding effect of income. The anthropometric indicators used were
expressed as percentages of the National Centre for Health Statistics median [16] and were
treated as dependent variables, with nutrition-knowledge score as the main effect and literacy of
mothers and per capita monthly income as covariates.

TABLE 1. Relation between mothers' literacy status and their children's nutritional status
as indicated by weight for age
Nutritional statuea
Child's age group
0-12 mo 13-36 mo 37-72 mo

N % N % N %

Illiterate
4th 24 2.7 38 2.7 23 1.6

3rd 92 10.3 197 14.0 179 12.2

2nd 203 22.7 484 34.4 502 34.3

1st 262 29.2 413 29.4 490 33.5

Normal 315 35.2 273 19.4 269 18.4

Literate
4th 1 2.2 3 2.9 0 0.0

3rd 1 2.2 9 8.6 10 1 1.1

2nd 8 17.8 30 28.6 26 28.9

1st 16 35.6 38 36.2 30 33.3

Normal 19 42.2 25 23.8 24 26.7

Chi-square
4.476b 5.587b 5.318b

a. "4th"- "1st" indicate degrees of malnutrition.

b. Not significant.

TABLE 2. Relation between mothers' literacy status and their children's nutritional status
as indicated by height for age

Nutritional statusa
Child's age group
0-12 mo 13-36 mo 37-72 mo
 N % N % N %

Illiterate
Severe 48 5.7 271 19.3 326 22.3

Moderate 161 18.0 397 28.3 397 27. 1

Mild 309 34.5 414 29.5 400 27.3

Normal 378 42.2 323 23.0 340 23.2

Literate
Severe 3 6.7 11 10.5 13 14.4

Moderate 2 4.4 33 31.4 25 27.8

Mild 15 33.3 38 36.2 32 35.6

Normal 25 55.6 23 21.9 20 22.2

Chi-square
6.464b 5.933b 4.494b

a. "Severe," "moderate," and "mild" indicate degrees of malnutrinon.

b. Not significant.

Results

Tables 1-3 show the agreement between the mothers' literacy status and the nutritional status of
their children by weight for age, height for age, and weight for height respectively. A positive
relationship appears to exist: for each of the three indicators, a higher percentage of the children
of the literate mothers were of normal status or had a mild grade of malnutrition than of the
children of the illiterate mothers, and a smaller proportion of the children of the literate mothers
were moderately or severely malnourished than of those of the illiterate mothers. The chi-square
test, however, indicated that these differences were not statistically significant.

The mothers' nutrition knowledge on the whole was found to be poor; their mean score was 1.6
out of the maximum of 9.
Tables 4-6 show the relations between the mothers' nutrition-knowledge scores and their
children's nutritional status by the three indicators. Good agreement is seen between the scores
and the children's weight for age and height for age for all except the oldest age group (37-72
months): a significantly higher percentage of the children 0-36 months old whose mothers scored
3 or less were moderately or severely malnourished than of those of the mothers with a score of 4
or more (tables 4 and 5). A similar effect is seen in the older group, but it is not statistically
significant. A significant relation is seen between the mothers' scores and wasting as assessed by
weight for height in the children in all the age groups (table 6).

As to the relationship between the mothers' nutrition knowledge and literacy, although the great
majority of both literate and illiterate mothers had scores in the poor range (1-3 points), a
significantly larger proportion of the literate group (20.9%) than of the illiterate group (4.6%)
had scores of 4 or more, while 15.2% of the illiterate group but only 4.8% of the literate group
had a score of 0 (table 7). Both literacy and nutrition-knowledge scores were significantly related
to family income : a higher proportion of illiterate mothers and those with scores of 0 or 1-3 were
from families below the poverty line (Rs 65 per capita per month) than their respective
counterparts.

Table 9 shows the results of the analysis of covariance used to determine the effect of the
mothers' nutrition knowledge on their children's nutritional status. Both literacy and income were
found to be significant intervening factors in the test for the effect of the nutrition-knowledge
score on weight for age, whereas only literacy was a significant factor in the case of height for
age and income in the case of weight for height. The combined effect of these factors was also
significant. Thus, the effect of the nutrition-knowledge scores on weight for age, height for age,
and weight for height was highly significant when controlled for literacy and family per capita
monthly income.

TABLE 3. Relation between mothers' literacy status and their children's nutritional status
as indicated by weight for height

Child's age group

Nutritional status 0-12 mo 13-36 mo 37-72 mo

N % N % N %

Illiterate
Severe 60 6.7 77 5.5 37 2.5

Moderate 149 16.6 345 24.6 288 19.7

Mild 273 30.5 608 43.3 711 48.6

Normal 414 46.2 375 26.7 427 29.2

Literate
Severe 2 4.4 4 3.8 2 2.2

Moderate 5 11.1 20 19.0 11 12.2

Mild 16 35 6 40 38 1 39 43.3

Normal 22 48.9 41 39.0 38 42.2

Chi-square
1.555a 7.760a 7.783a

a. Not significant.

TABLE 4. Relation between mothers' nutrition-knowledge scores and their children's

nutritional status as indicated by weight for age

Child's age group

Nutritional status 0- 12 mo 13-36 mo 37-72 mo

N % N % N %

Score 0
4th 5 4.2 16 6.5 5 2.3

3rd 31 26.3 59 23.8 32 14.4

2nd 41 34.7 86 34.7 84 37.8

1st 24 20.3 61 24.6 69 31.1

Normal 17 14.4 26 10.5 32 14.4

Score 1-3
4th 18 2.4 23 1.9 18 1.5
3rd 60 7.9 142 12.0 151 12.2

2nd 157 20.6 409 34.5 417 33.7

1st 234 30.7 362 30.6 417 33.7

Normal 293 38.5 248 20.9 235 19.0

Score 3-4
4th 2 3.3 2 2.6 0 0

3rd 2 3.3 5 6.4 6 6.5

2nd 13 21.3 19 24.4 27 29.0

1st 20 32.8 28 35.9 34 36.6

Normal 24 39.3 24 30 8 26 28.0

Chi-square
69.334**** 61.599**** 14.441

****P<.001.

TABLE 5. Relation between mothers' nutrition-knowledge scores and their children's

nutritional status as indicated hv height for age

Child's age group

Nutritional status 0-12 mo 13-36 mo 37-72 mo

N % N % N %

Score 0
Severe 13 11.0 66 26.6 50 22.5

Moderate 31 26.3 57 23.0 57 25.7

Mild 30 25.4 72 29.0 67 30.2

Normal 44 37.3 53 21.4 48 21.8

Score 1-3
Severe 37 49 205 17.3 276 22.3

Moderate 124 16.3 354 29.9 341 27.5

Mild 269 35.3 354 29.9 335 27.3

Normal 332 43.6 271 22.9 283 22.9

Score  4
Severe 1 1.6 11 14.1 13 14.0

Moderate 8 13.1 19 24.4 24 25.8

Mild 25 41.0 26 33.3 27 29.0

Normal 27 44.3 22 28.2 29 31.2

Chi-square
20.122*** 15.927** 6.577

**P<.025.
***P<.005.

TABLE 6. Relation between mothers' nutrition-knowledge scores and their children's

nutritional status as indicated by weight for height

Child's age group

Nutritional status 0-12 mo 13-36 mo 37-72 mo

N % N % N %

Score 0
Severe 18 15.3 26 10.5 8 3.6

Moderate 23 19.5 94 37.9 61 27.5

Mild 38 32.2 89 35.9 95 42.8

Normal 39 33.1 39 15.7 58 26.1

Score 1-3
Severe 41 5.4 55 4.6 31 2.5

Moderate 124 16.3 256 21.6 225 18.2

Mild 228 29.9 528 44.6 613 49.5

Normal 269 48.4 345 29.1 369 29.8

Score  4
Severe 3 4.9 0 0.0 0 0.0

Moderate 7 11.5 15 19.2 13 14.0

Mild 23 37.7 31 39.7 42 45.2

Normal 28 45.9 32 41.0 38 40.9

Chi-square
23.422**** 63.388**** 20.034***

***P<.005.
****P<.001.

TABLE 7. Agreement between mothers' literacy status and nutrition-knowledge scores

Score Illiterate Literate

N % N %

0 379 15.2 8 4.8

1-3 2,003 80.2 124 74.3

>=4 116 4.6 35 20.9

Chi-square
104.24*

****P<.001.

TABLE 8. Relation between family per capita monthly income and mothers literacy and
nutritionknowledge scores

Literacy Score
Income
Illiterate Literate 0 1-3 4
N % N % N % N % N %

 Rs 65 2,972 79.0 135 56.3 488 83.0 2,439 76.6 180 77.6

> Rs 65 792 21.0 105 43.8 100 17.0 745 23.4 52 22.4

Chi-square
65 627* * * * 11.664* * *

*** P<.005.
**** P<.001.

TABLE 9. Analysis of covariance of mothers' nutrition-knowledge scores and their

children's nutritional-status indicators

F value
Wt/age Ht/age Wt/ht

Nutrition knowledge
46.915* 11.385* 33. 577*
score

Covariates 8.420* 5.604* 3.811*

literacy 9.546* 1.018 7.471 *

income 4.138* 8.577* 0.851

* P <.05.

Discussion

Our data reveal a definite but non-significant association between the literacy of the mothers and
the nutritional status of their children. However, it may be noted that statistical significance is a
strong function of, among other things, the number of observations (sample size), which in the
present case is extremely skewed, with only 6% of the women in the literate category.

Other investigators [4, 17] have shown that the number of years of education of mothers had a
definite relationship with the proportion of malnourished children and related this to the fact that
children's mean daily intake of nutrients increased with the increase in the mothers' educational
level. Also, the intake of nutrients by all the children whose mothers had an education only up to
primary and middle grades was far below the recommended allowances. In the present study.
although the mothers' education level was not considered because of their very low rate of
literacy, the difference in the weight and height of children of literate and illiterate mothers was
clear.

The relationship between the nutrition knowledge of the mothers and the nutritional status of
their children was much stronger. One study [8] reported that mothers' nutrition-knowledge
scores were associated with the long-term well-being of children represented by height for age.
Our findings show that the mothers' nutrition knowledge did not affect the weight for age and
height for age of the children 37-72 months old, whereas acute malnutrition as indicated by
weight for height was significantly related to the mothers' knowledge in children of all age
groups. It is likely that for older children, other factors have a stronger influence on nutritional
status than mothers' nutrition knowledge. Also, by the time children are three or four years old,
they may have younger siblings who require maternal care and attention, in which case the older
children are likely to be ignored.

Some studies have found no relationship of mothers' nutrition knowledge on the nutritional status
of children. On the basis of arbitrarily prepared knowledge and belief scores, one such study
reported that the mothers of well-nourished children were as ignorant about essential facts
regarding nutrition as those of undernourished children [9]. Similarly, another found that
maternal comprehension of home based growth charts had no effect on children's growth [10].

Our findings show a significant difference between the nutrition-knowledge scores of literate and
illiterate mothers, with the former having better scores. Income was also significantly positively
related to both maternal literacy and nutrition knowledge. The cause-effect relationship in this
instance appears to be bi-directional, in that higher income reflects better maternal literacy status
and nutrition knowledge and vice versa. Victora et al. [5] demonstrated that income and parental
education are strongly correlated, and that maternal education affects child nutritional status even
when family income is taken into consideration.

The present study also shows that maternal literacy and nutrition knowledge exerted a significant
intervening effect on at most two of the selected anthropometric indicators. However, where
significant, regardless of confounding effects. the nutrition-knowledge score had a significant
positive effect on all the indicators of child nutritional status. The evidence for a causal
association is strongest when it remains statistically significant after family income is taken into
consideration. Thus, if mothers have sufficient nutrition knowledge, it is effective in improving
the nutritional status of their children. This implies that, although all women do need formal
education, nutrition education is a short-term intervention that will have a considerable impact on
the community. The need for such education for women is therefore urgent and great in rural and
tribal India.

References

1. Caldwell JC. Maternal education as a factor in child mortality. World Health Forum 1981;2:75-78.
2. Chen LC. Primary health care in developing countries: overcoming operational, technical and
social barriers. Lancet 1986;2:1260-65.

3. Bhuiya A, Zimicki S, D'Sonza S. Socioeconomic differentials in child nutrition and morbidity in a

rural area of Bangladesh. J Trop Pediatr 1986;32:17-23.

4. Bhuiya A, Wojtyniak B, D'Souza S, Zimicki S. Socio-economic determinants of child nutritional

status: boys versus girls. Food Nutr Bull 1986;8(3):3-7.

5. Victora CG, Vaughan PJ, Kirkwood BR, Martinez JC, Barcelos LB. Risk factors for malnutrition in
Brazilian children: the role of social and environmental variables. Bull WHO 1986;64:299-309.

6. Kimati VP. Who is ignorant? rural mothers who feed their well nourished children or the
nutrition experts? the Tanzanian story. J Trop Pediatr 1986;32:130-36.

7. Srikantia SG, Sastry CY. Effects of maternal attributes on malnutrition in children. Proceedings of
the First Asian Congress of Nutrition. Hyderabad, India. 1972:584

8. Smith MF, Paulsen SK, Fougers W, Ritchey SJ. Socioeconomic, education and health factors
influencing growth of rural Haitian children. Ecol Food Nutr 1983;13:99-lOX.

9. Walia BNS, Gambhir SK. The relationship between childhood malnutrition and maternal
knowledge and beliefs regarding nutrition. Ind Pediatr 1975; 12: 563-68.

10. Grant K, Stone T. Maternal comprehension of home-based growth charts and its effect on
growth. J Trop Pediatr 1986;32:255-57.

11. Bairagi R. Dynamics of child nutrition in rural Bangladesh. Ecol Food Nutr 1982:13:173-78.

12. Gujarat Directorate of Health, Medical Services and Medical Education. Health statistics.
Ahmedabad, Gujarat, India, 1984.

13. Jelliffe DB. The assessment of nutritional status of the community. WHO Monograph Series, no.
53. Geneva: World Health Organization, 1966.

14. Indian Academy of Pediatrics. Classification suggested by the Nutrition Subcommittee of the
Indian Academy of Pediatrics, 1971-72. Ind Pediatr 1972;6:360.
 15. Waterlow JC. Medical practice. Br Med J 1972;3:566-69.

16. National Center for Health Statistics. NCHS growth curves for children: birth-18 years. United
States. DHEW publication 78-1650. Washington, D.C.: Department of Health, Education, and
Welfare. 1977.

17. Devdas RP, Rajlakshmi R. Kaveri R. Influence of family income and parents' education on the
nutritional status of preschool children. Ind J Nutr Dietet 1980; 17:237-43.

http://www.unu.edu/unupress/food/8F104e/8F104E06.htm