Sant-Rayn Pasricha, James Black, Sumithra Muthayya, Anita Shet, Vijay Bhat,

Determinants of Anemia Among Young Children in Rural India
Sant-Rayn Pasricha, James Black, Sumithra Muthayya, Anita Shet, Vijay Bhat,
Savitha Nagaraj, N. S. Prashanth, H. Sudarshan, Beverley-Ann Biggs and Arun S.
Pediatrics 2010;126;e140; originally published online June 14, 2010;
DOI: 10.1542/peds.2009-3108
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly
publication, it has been published continuously since 1948. PEDIATRICS is owned,
published, and trademarked by the American Academy of Pediatrics, 141 Northwest Point
Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2010 by the American Academy
of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.
Downloaded from by guest on August 22, 2014
Determinants of Anemia Among Young Children in
Rural India
WHAT’S KNOWN ON THIS SUBJECT: The immense burden of
anemia among toddlers in developing countries, particularly
India, has been documented and attributed to iron deficiency.
Limited data are available, however, regarding the biological,
nutritional, and socioeconomic etiologies of anemia, especially in
rural settings in which the prevalence is maximal.
WHAT THIS STUDY ADDS: We present a comprehensive
evaluation in rural Indian children of biological (micronutrient,
infectious disease, and genetic), maternal, and socioeconomic
factors possibly associated with hemoglobin. In addition to iron
status, folate level, maternal hemoglobin level, family wealth/food
insecurity, and hemoglobinopathy were also independently
associated with hemoglobin.
AUTHORS: Sant-Rayn Pasricha, MBBS, MPH,a,b James
Black, MBBS, PhD, FAFPHM,a Sumithra Muthayya, BSc,
PhD,b Anita Shet, MBBS, MD, FAAP,c,d Vijay Bhat, MSc,e
Savitha Nagaraj, MBBS, MD,f N. S. Prashanth, MBBS,
MPH,g H. Sudarshan, MBBS,g Beverley-Ann Biggs, MBBS,
PhD, FRACP,h and Arun S. Shet, MBBS, MDd,i
aNossal Institute for Global Health, University of Melbourne,
Parkville, Victoria, Australia; bHematology and Nutrition Units
and Departments of cPediatrics, fMicrobiology, and iMedical
Oncology, St Johns National Academy of Health Sciences,
Bangalore, India; eBiochemistry Laboratory, Manipal Hospital,
Bangalore, India; gKaruna Trust, Bangalore, India; hDepartment
of Medicine and Victorian Infectious Diseases Service, University
of Melbourne, Royal Melbourne Hospital, Parkville, Victoria,
Australia; and dDivision of Global Health, Department of Public
Health Sciences, Karolinska Institutet, Stockholm, Sweden
anemia, India, child preschool, iron-deficiency anemia, public
health, poverty, food security
OBJECTIVE: More than 75% of Indian toddlers are anemic. Data on
factors associated with anemia in India are limited. The objective of
this study was to determine biological, nutritional, and socioeconomic
risk factors for anemia in this vulnerable age group.
METHODS: We conducted a cross-sectional study of children aged 12
to 23 months in 2 rural districts of Karnataka, India. Children were
excluded if they were unwell or had received a blood transfusion. Hemoglobin, ferritin, folate, vitamin B12, retinol-binding protein, and
C-reactive protein (CRP) levels were determined. Children were also
tested for hemoglobinopathy, malaria infection, and hookworm infestation. Anthropometric measurements, nutritional intake, family
wealth, and food security were recorded. In addition, maternal hemoglobin level was measured.
RESULTS: Anemia (hemoglobin level ⬍ 11.0 g/dL) was detected in
75.3% of the 401 children sampled. Anemia was associated with iron
deficiency (low ferritin level), maternal anemia, and food insecurity.
Children’s ferritin levels were directly associated with their iron intake
and CRP levels and with maternal hemoglobin level and inversely associated with continued breastfeeding and the child’s energy intake. A
multivariate model for the child’s hemoglobin level revealed associations with log(ferritin level) (coefficient: 1.20; P ⬍ .001), folate level
(0.05; P ⬍ .01), maternal hemoglobin level (0.16; P ⬍ .001), family
wealth index (0.02; P ⬍ .05), child’s age (0.05 per month; P ⬍ .005),
hemoglobinopathy (⫺1.51; P ⬍ .001), CRP level (⫺0.18; P ⬍ .001), and
male gender (⫺0.38; P ⬍ .05). Wealth index and food insecurity could
be interchanged in this model.
CONCLUSIONS: Hemoglobin level was primarily associated with iron
status in these Indian toddlers; however, maternal hemoglobin level,
family wealth, and food insecurity were also important factors. Strategies for minimizing childhood anemia must include optimized iron
intake but should simultaneously address maternal anemia, poverty,
and food insecurity. Pediatrics 2010;126:e140–e149
WHO—World Health Organization
NFHS—National Family Health Survey
PHC—primary health center
INR—Indian rupees
RBP—retinol-binding protein
CRP—C-reactive protein
CI— confidence interval
Accepted for publication Mar 26, 2010
Address correspondence to Arun S. Shet, MBBS, MD, Department
of Medical Oncology, St Johns National Academy of Health
Sciences, Sarjapur Road, Bangalore 560034, India. E-mail:
[email protected]
PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).
Copyright © 2010 by the American Academy of Pediatrics
FINANCIAL DISCLOSURE: The authors have indicated they have
no financial relationships relevant to this article to disclose.
Downloaded from by guest on August 22, 2014
The World Health Organization (WHO)
has estimated that, globally, 1.62 billion people are anemic, with the highest prevalence of anemia (47.4%)
among preschool-aged children; of
these 293 million children, 89 million
live in India.1 The third National Family
Health Survey (NFHS) 2005–2006 revealed that at least 80% of Indian children aged 12 to 23 months were anemic.2 Anemia was especially prevalent
among rural children,2 and the majority of India’s population (72.2%) is rural.3 However, despite recent economic development4 and the existence
of a national anemia-control program,5
the prevalence of anemia in India between 2000 and 2005 increased from
75.3% to 80.9% in children aged 6 to 36
months.2,6 Alleviating childhood irondeficiency anemia is a public-health
priority, because anemia is associated
with impaired cognitive and psychomotor development.7,8
Iron deficiency is believed to be the
most important cause of anemia
among children in India9 and is attributable to poor nutritional iron intake
and low iron bioavailability.10 Other
factors, including folate and vitamin
B12 and A deficiencies, malaria infection, hookworm infestation, and hemoglobinopathies, are also associated
with childhood anemia.11–13 To our
knowledge, no previous report in the
published literature has described the
relative contribution of these factors
to anemia in rural Indian children. To
effectively control this problem, health
care providers must have a comprehensive understanding of the etiologic
factors associated with anemia.
We hypothesized that low hemoglobin
concentrations in rural Indian children
primarily result from micronutrient
(especially iron) deficiencies attributable to poor nutritional intake compounded by adverse socioeconomic
conditions and food insecurity. To test
this hypothesis we conducted laboraPEDIATRICS Volume 126, Number 1, July 2010
tory, nutritional, anthropometric, and
socioeconomic evaluations in a crosssection of rural Indian children aged
12 to 23 months.
Study Site and Participants
Study participants were members of eligible sample populations in villages
served by 2 primary health centers
(PHCs), the basic units of health care delivery in rural India. The Gumballi PHC in
the Chamarajnagar district, 112 miles
south of Bangalore, serves 21 700 people
in 13 villages. The Sugganahalli PHC in
the Ramnagara district, 56 miles northwest of Bangalore, serves 14 400 people
in ⬃80 villages.14 Both districts have
agrarian economies and per-capita annual incomes (Chamarajnagar, Indian
rupees [INR] 22 006 [US $478]; Ramnagara, INR 26 009 [US $565]) that reflect
the state and national averages (Karnataka overall, INR 26 123 [US $567]; India
overall, INR 25 825 [US $561]).15
We randomly selected 3 of 4 subcenters
of each PHC.16 Lists of children living in
the villages were compiled from lists obtained from both PHCs and from Anganwadi (child care centers for preschoolaged children), and the information was
confirmed by investigators who conducted house-to-house visits. All children
aged 12 to 23 months in selected villages
were eligible, unless the child was
acutely unwell or had received a blood
transfusion. The detailed study methods
have been published.17
Study Procedures
The food-security questionnaire module
was adapted from the Household Food
Insecurity Access Scale, which has been
validated for use in settings in developing countries.18 The scale covers perceptions about food insecurity and enables
calculation of a score from 0 (no food
insecurity) to 27 (maximum food insecurity).19 The Wealth Index, an estimation of
household wealth in which assets are as-
signed a weighted score (maximum of
63), 20 was adapted from the third NFHS.2
Study participants used containers of
standardized sizes for a 24-hour dietary
recall to estimate nutrient intake,21
which was expressed as a percentage of
the Indian recommended daily intake.22
Information on continued and previous
breastfeeding practices was obtained
with specific questions about whether
the child was currently breastfeeding,
the duration of exclusive breastfeeding,
and the child’s age at introduction of
complementary foods and age at breastfeeding cessation.
Each child’s length (from the crown of
the head to the heel) (Seca 210 [Seca,
Hamburg, Germany) and weight (Seca
872) were measured. Height for age,
weight for age, and weight for length
were calculated. For these variables,
results with z scores less than ⫺2
were defined as stunting, underweight, and wasting, respectively, in
accordance with the 2008 WHO child
growth standards.23
Mothers underwent field estimation of
capillary hemoglobin level (HemoCue
201⫹ [HemoCue, Angelholm, Sweden]). Venous blood (3 mL) was drawn
from each child, processed appropriately in the PHC laboratory within 6
hours of collection, and then packed
with ice and transported to the reference laboratory. Samples were analyzed within 48 hours of collection.
Laboratory Assays
Laboratory assays were performed as
follows: automated complete blood examination (Sysmex XT-2000i [Sysmex Inc,
Kobe, Japan]) (used only for hemoglobin
estimation); serum ferritin, folate, and vitamin B12 (electrochemiluminescent immunoassay, ELECSYS 2010 [ELECSYS, Hitachi High Technologies Corporation,
Tokyo, Japan]; reagents from Roche Diagnostics [Penzberg, Germany]), retinolbinding protein (RBP), high-sensitivity Creactive protein (CRP) (nephelometry,
Downloaded from by guest on August 22, 2014
Siemens BN-Prospec Nephelometer [Siemens, Marburg, Germany]), and hemoglobin variant (high-performance liquid
chromatography, Biorad D10 [Biorad
Laboratories Inc, Hercules, CA]). Thick
and thin blood films were prepared by
use of the Jaswant-Singh-Bhattacherji
method and evaluated for malaria parasites by technicians for the National Malaria Control Program. For study participants for whom stool samples were
returned (n ⫽ 142), stool was evaluated
microscopically for hookworm ova by
use of wet mounts.24
Total number of children
potentially eligible in
target villages
Total number of children
presenting for study;
examined for eligibility
Total number of children
included, questionnaires
Previous blood
transfusion 3
Presented with
Anemia was defined as a hemoglobin
level of ⬍11 g/dL in children, ⬍12 g/dL in
nonpregnant women, and ⬍11 g/dL in
pregnant women, on the basis of WHO
definitions.9 Iron deficiency was defined
as a ferritin level of ⬍12 ng/mL, or ⬍30
ng/mL if the CRP level was ⬎5 mg/L.9 Using the manufacturer’s reference
ranges, we defined vitamin B12 deficiency as serum vitamin B12 level of
⬍210 pg/mL and folate deficiency as a
serum folate level of ⬍3.3 ng/mL. Biochemical evidence of inflammation was
defined as a CRP level of ⬎5 mg/L25 and
␤-thalassemia trait as a hemoglobin A2
level of ⬎3.5%.26 Although RBP level is
highly correlated with serum retinol levels, reference ranges in the pediatric
population are unclear and cutoffs were
not applied.27,28 Malaria was diagnosed if
plasmodium parasites (trophozoites,
schizonts, or gametocytes) were identified in serum. Results for hookworm ova
in stool samples were expressed as
present or absent.
Ethics Considerations
Information obtained through community consultation was used to formulate the study design and procedures.
Plain-language statements explaining
the study were provided to and written
informed consent was obtained from
the guardians of all child participants.
The study was approved by the ethics
Exclusions 10
Febrile 7
Away from
Blood samples collected
and analyzedb
401 (EDTA)
396 (serum)
Subject enrollment.17 a Compiled from Anganwadi Centre lists, health worker’s lists, and house-tohouse surveys (when lists were inadequate). b Blood could not be collected from 4 children, and
serum was not collected from another 5 subjects.
committees of St John’s National Academy of Health Sciences, Bangalore, India, and the Faculty of Medicine, Dentistry and Health Sciences, University
of Melbourne, Australia.
Statistical Methods
Data were entered into Epi Info 3.4.3 (US
Centers for Disease Control and Prevention, Atlanta, GA) and exported to Stata 9
(Stata Corp, College Station, TX) for analysis. A sample size of 390 ensured a ⫾5%
range for the 95% confidence intervals
(CIs) of estimates of prevalence. The
study had 80% power to detect regression coefficients between continuous
variables for which the coefficient was at
least 0.1 and the correlation between
variables was at least 0.2. Linear regression performed by using continuous
variables retained maximum information29; in particular, hemoglobin level
was analyzed (rather than “anemia”) as
the outcome variable, because the
threshold for defining anemia has age
and ethnic ambiguities.30 CRP level was
analyzed as an ordered categorical variable. Associations between risk factors
and outcomes were first evaluated by using univariate linear regression. A
multiple-regression model was then iteratively developed. Variables were retained if the P value for their coefficient
remained ⬍.05. Standardized (␤) coefficients (coefficient standardized with a
mean of 0 and SD of 1) were calculated.
An R2 value was used to determine the
variation in hemoglobin level revealed by
the model. The Shapiro-Wilk test was
used to determine if residual values for
the model had a normal distribution.
Between August and October 2008,
88.3% of 470 eligible children living in
the selected villages were recruited
(Fig 1).17 Mean values with 95% CIs in
Downloaded from by guest on August 22, 2014
parenthesis are shown unless otherwise indicated.
TABLE 1 Mean Nutrient Intake From Non–Breast Milk Sources During the Previous 24 Hours
Demographics and Food Insecurity
Iron, mg/24 h
Folate, mg/24 h
Vitamin B12, mg/24 h
Vitamin A, mg/24 h
Energy, kcal/24 h
The mean age of children was 17.2
months, and 204 of 405 (50.3%) were
boys. Mean maternal age was 23.2
years, and 45 of 376 mothers (12%)
were pregnant. Overall socioeconomic
status was low, with a mean wealth index of 18 (17.0 –19.0). The median food
insecurity score was 3 (0 –27), and
more than half of the mothers (212 of
402 [52.7%]) reported some degree of
household food insecurity during the
previous month. Food insecurity and
wealth indices were inversely associated (Spearman rank correlation:
⫺0.65; P ⬍ .001).
Nutritional Intake
The mean nutritional intake for children is shown in Table 1. Mean iron
intake from non– breast milk sources
consumed during the previous 24
hours was 1.4 (1.3–1.5) mg, which was
11.2% of the recommended daily intake for Indian children. Children who
were still breastfed had lower iron intake from complementary food (1.10
[1.00 –1.20] mg) than children who
were no longer breastfed (1.99 [1.85–
2.13] mg; P ⬍ .001). Iron intake was
positively associated with wealth index
(coefficient: 0.01 [0.00 – 0.02]; P ⬍
.005) but not with food insecurity.
Mean growth indices are shown in Table 2. Almost one-third of the children
were underweight (129 of 400 [32.3%]
[range: 27.7%–36.9%]). Stunting was
seen in 115 of 401 children (28.7%
[24.2%–33.1%]), and wasting was seen
in 83 of 400 (20.8% [16.8%–24.7%]) (Table 3). Length-for-age z scores between
boys and girls were similar, but absolute length was greater in boys (77.8 vs
75.6 cm; P ⬍ .001).
PEDIATRICS Volume 126, Number 1, July 2010
Mean (95% CI)
Mean Intake,
% Consuming
⬍75% RDI
1.4 (1.3–1.5)a
33.8 (31.4–36.3)a
0.31 (0.30–0.36)a
131.0 (113.1–151.8)b
415.0 (393.9–436.1)c
27.2 (22.7–31.7)
46.6 (41.6–51.6)
65.5 (60.8–70.3)
98.5 (97.2–99.7)
CI indicates confidence interval; RDI, recommended daily intake.
a Geometric mean.
b Geometric mean; transformation after addition of 1; 1 subtracted after exponentiation.
c Arithmetic mean.
TABLE 2 Mean Indices of Hemoglobin Level and Associated Factors
Mean (95% CI)
Hemoglobin (n ⫽ 401), g/dLa
Ferritin (n ⫽ 396), ng/mLb
Vitamin B12 (n ⫽ 396), pg/mLb
Folate (n ⫽ 396), ng/mLa
RBP (n ⫽ 382), g/Lb
CRP (n ⫽ 396), mg/mLc
Length for age (n ⫽ 401), z scorea
Weight for age (n ⫽ 400), z scorea
Weight for length (n ⫽ 400), z scorea
9.75 (9.59 to 9.91)
10.97 (10.09 to 11.92)
420.1 (403.0 to 437.9)
9.88 (9.54 to 10.22)
0.028 (0.026 to 0.030)
0.91 (0.77 to 1.06)
⫺1.45 (⫺1.57 to ⫺1.34)
⫺1.55 (⫺1.65 to ⫺1.45)
⫺1.14 (⫺1.24 to ⫺1.04)
CI indicates confidence interval.
a Arithmetic mean.
b Geometric mean.
c Geometric mean; transformation following addition of 1; 1 subtracted after reverse transformation.
Anemia and Associated Conditions
Mean hemoglobin levels and biological
factors possibly associated with hemoglobin levels are shown in Table 2.
Anemia was detected in 75.3% of the
children (hemoglobin level: 9.75 g/dL
[9.59 –9.91]). Anemia was prevalent to
a similar extent in nonpregnant
women (63.3%; hemoglobin level: 11.2
g/dL [11.0 –11.4]) and pregnant mothers (61.0%; hemoglobin level: 10.6 g/dL
[10.1–11.1]). Childhood iron deficiency
(ferritin level: ⬍12 or ⬍30 ng/mL if
CRP level was ⬎5 mg/L) was seen in
61.9% (57.1%– 66.7%). Children with
anemia were more likely to have iron
deficiency (odds ratio [OR]: 6.1; P ⬍
.001), maternal anemia (OR: 1.9; P ⬍
.01); continued breastfeeding (OR: 1.6;
P ⬍ .05), and food insecurity (OR: 2.2;
P ⬍ .005) compared with children
without anemia (Table 3).
Associations With Hemoglobin
Hemoglobin was positively associated
with child’s age (coefficient: 0.05
[0.00 – 0.10]; P ⬍ .05) and was lower in
boys compared with girls (t test, mean
difference in hemoglobin: ⫺0.51 g/dL
[⫺0.19 to ⫺0.82]; P ⬍ .01). Univariate
regression analysis, controlled for age
and gender, revealed that hemoglobin
level was positively associated with
ferritin, folate, and vitamin A intake;
wealth; and maternal hemoglobin level
and negatively associated with food insecurity (Table 4). Multiple regression
analysis results indicated that children’s hemoglobin levels were primarily associated with ferritin intake but
also positively associated with maternal hemoglobin level, child folate intake and age, and family wealth and
were inversely associated with male
gender, CRP level, and presence of the
␤-thalassemia trait (Table 5). The foodinsecurity score could be substituted
for wealth index without affecting
other factors. The residuals for this
model were normally distributed
(Shapiro-Wilk test, P ⬎ .05), and the R2
value was 0.51.
Downloaded from by guest on August 22, 2014
TABLE 3 Proportion of Children With Anemia (Hemoglobin ⬍ 11 g/dL) and Associated Conditions
Laboratory indices
Iron deficiencya
Folate deficiencyb
Vitamin B12 deficiencyc
Hookworm infestationf
Malaria infectionf
Nutritional intake
Low iron intakeg
Continued breastfeedingh
Food insecurityi
Maternal hemoglobin level: maternal anemiaj
Child growth
% (95% CI)
Anemic, %
(95% CI)
Not Anemic, %
(95% CI)
Odds Ratio
61.9 (57.1 to 66.7)
1.3 (0.2 to 2.4)
2.8 (1.2 to 4.4)
9.9 (6.9 to 12.8)
1.3 (0.2 to 2.3)
13.4 (7.7 to 19.0)
72.2 (67.1 to 77.3)
1.0 (⫺0.1 to 2.1)
2.7 (0.1 to 4.5)
10.7 (7.2 to 14.2)
1.7 (0.2 to 3.1)
16.2 (8.9 to 23.4)
29.9 (20.6 to 39.2)
2.1 (⫺0.8 to 4.9)
3.1 (⫺0.0 to 6.5)
7.2 (2.1 to 12.4)
7.5 (⫺0.7 to 15.7)
6.1 (3.6 to 10.5)
0.5 (0.1 to 5.9)
0.9 (0.2 to 5.1)
1.5 (0.6 to 4.3)
2.4 (0.6 to 13.4)
57.9 (53.1 to 62.8)
52.7 (47.8 to 57.6)
63.1 (58.1 to 68.1)
60.9 (55.4 to 66.4)
57.1 (51.5 to 62.8)
66.8 (61.2 to 72.4)
49.0 (39.1 to 58.9)
38.1 (28.3 to 48.0)
51.1 (40.8 to 61.6)
1.6 (1.0 to 2.6)
2.2 (1.3 to 3.6)
1.9 (1.1 to 3.2)
28.7 (24.2 to 33.1)
32.3 (27.7 to 36.9)
20.8 (16.8 to 24.7)
29.2 (24.1 to 34.4)
32.6 (27.2 to 37.9)
19.9 (15.3 to 24.4)
26.5 (17.6 to 35.4)
30.9 (21.6 to 40.3)
22.9 (14.4 to 31.5)
1.1 (0.7 to 2.0)
1.1 (0.6 to 1.8)
0.8 (0.5 to 1.5)
CI indicates confidence interval.
a Ferritin level ⬍ 12 ng/mL, or ⬍ 30 ng/mL if CRP ⬎ 5 mg/L.
b Serum folate level ⬍ 3.3 ng/mL.
c Serum vitamin B level ⬍ 210 pg/mL.
d CRP level ⬎ 5 mg/L.
e Hemoglobin A ⬎ 3.5%.
f Parasites identified on microscopy.
g Iron intake ⬎75% of Indian recommended daily intake.
h The child was continuing to receive breast milk at the time of the study.
i Household Food Insecurity Access Scale score ⱖ1.
j Hemoglobin level ⬍ 12 g/dL, or hemoglobin level ⬍ 11 g/dL if pregnant.
k Length-for-age z score less than ⫺2.
l Weight-for-age z score less than ⫺2.
m Weight-for-length z score less than ⫺2.
Associations With Ferritin
By using univariate regression we
found that child’s (log)ferritin level
was associated with maternal hemoglobin level (coefficient: 0.09 [0.03–
0.13]; P ⬍ .01) and CRP level (coefficient: 0.18 [0.12– 0.25]; P ⬍ .001) but
not with child’s age, family wealth,
food insecurity, or child’s nutrient intake. According to multiple regression
analysis results, child’s log(ferritin)
was independently, positively associated with maternal hemoglobin level
(coefficient: 0.07 [0.02– 0.12]; P ⬍ .01),
CRP level (coefficient: 0.20 [0.13– 0.27];
P ⬍ .01), and log(iron intake) (coefficient: 0.25 [0.01– 0.49]; P ⬍ .05). On the
other hand, child’s (log)ferritin was
negatively associated with a history of
breastfeeding beyond 12 months (coefficient: ⫺0.24 [⫺0.42 to ⫺0.06]; P ⬍
.01) and increased calorie intake from
complementary foods (coefficient:
⫺0.001 [⫺0.002 to ⫺0.000]; P ⬍ .02).
childhood anemia in rural Indian toddlers and raise questions regarding
anemia-control policies.
The association between child’s hemoglobin level and child’s iron status and
maternal hemoglobin level may have
multiple pathways (Fig 2). For instance, antenatal anemia contributes
to low birth weight and prematurity,
both of which increase the risk of
childhood anemia.31 Severe maternal
anemia may also reduce breast milk
iron content.32 Children’s iron intake in
this population is universally low, particularly in children continuing to
breastfeed. Calories are predominantly available in cereals, which contain inhibitors of iron absorption.33
Thus, increased caloric intake may be
associated with reduced dietary-iron
bioavailability.34 Finally, the mother
and child share a socioeconomic envi-
We made the following observations
among 12- to 23-month-old rural Indian children. (1) Hemoglobin levels
in children were primarily related to
iron stores. (2) Levels of hemoglobin
were also associated with levels of
folate, CRP, and the ␤-thalassemia
trait. (3) Hemoglobin levels were independently associated with maternal hemoglobin level, family wealth,
and food insecurity. (4) Ferritin level
was positively associated with
dietary-iron intake and inversely
associated with continued breastfeeding beyond 1 year of age and increased energy intake from complementary foods. Taken together, these
data identify major determinants of
Downloaded from by guest on August 22, 2014
TABLE 4 Regression Coefficients (Univariate) Between Hemoglobin Level and Conditions
in the Child
Coefficient (95% CI)b
Proximate factor
Vitamin B12a
␤ -Thalassemia traitc
Hookworm ova in stool
Length for age, z score
Weight for age, z score
Weight for length, z score
Distal factor
Food insecuritye
Score 0
Score 1–5
Score 6–10
Score 11–15
Score ⱖ16
Maternal education
Maternal literacy
Maternal hemoglobin level
Child’s birth order
Iron intake
Folate intake
Vitamin B12 intake
Vitamin A intake
Child still breastfeedingf
1.17 (1.02 to 1.32)
0.08 (0.03 to 0.12)
0.07 (⫺0.31 to 0.45)
0.17 (⫺0.12 to 0.45)
⫺0.76 (⫺2.20 to 0.68)
0.01 (⫺0.12 to 0.13)
⫺0.69 (⫺1.49 to 0.11)
0.05 (⫺0.09 to 0.19)
0.04 (⫺0.11 to 0.20)
0.02 (⫺0.13 to 0.17)
0.02 (0.01 to 0.04)
⫺0.08 (⫺0.70 to 0.53)
⫺0.66 (⫺1.02 to ⫺0.30)
⫺0.94 (⫺1.68 to ⫺0.19)
⫺0.24 (⫺0.80 to 0.33)
0.10 (⫺0.01 to 0.22)
0.18 (⫺0.17 to 0.54)
0.28 (0.18 to 0.37)
⫺0.11 (⫺0.32 to 0.09)
0.16 (⫺0.07 to 0.43)
0.00 (⫺0.00 to 0.01)
0.37 (⫺0.14 to 0.87)
0.17 (0.05 to 0.27)
⫺0.28 (⫺0.62 to 0.06)
The regression coefficients were controlled for age and gender.
a Logarithmically transformed.
b The child’s age (in months) and gender were included in each regression equation but not reported separately; age and
gender were associated with hemoglobin in all equations.
c A positive result for ␤ -thalassaemia minor was coded as 1; all other results were coded as 0.
d The results for CRP level were categorized as 0, 0.1–1, 1–3, 3–5, 5–10, and ⬎10 mg/L.
e The n values for the food insecurity scores were 190 (score 0), 30 (score 1–5), 123 (score 6 –10), 21 (score 11–15), and 49
(score ⱖ 16).
f The results were coded as 1 if the child was still breastfeeding or 0 if the child was fully weaned.
TABLE 5 Multiple Regression Model of Associations With Hemoglobin
Association With
Coefficient (95% CI)
Log(ferritin), ng/mLb
Serum folate, ng/mL
CRP, mg/Lc
␤ -Thalassaemia traitd
Maternal hemoglobin, g/dL
Age of child, mo
1.20 (1.06 to 1.35)
0.05 (0.01 to 0.09)
⫺0.18 (1.06 to 1.35)
⫺1.51 (⫺2.53 to ⫺0.48)
0.16 (0.08 to 0.23)
0.02 (0.00 to 0.03)
⫺0.38 (⫺0.62 to ⫺0.13)
0.05 (0.02 to 0.09)
The regression coefficient was standardized with a mean of 0 and an SD of 1.
Logarithmically transformed.
c The results for CRP level were categorized as 0, 0.1–1, 1–3, 3–5, 5–10, and ⬎10 mg/L.
d A positive result for ␤ -thalassaemia minor was coded as 1; all other results were coded as 0.
e Gender was coded as 1 (male) or 0 (female).
ronment, and by the time the child is 12
months old, his or her dietary quality
may be similar.
PEDIATRICS Volume 126, Number 1, July 2010
Although growth may contribute to the
development of anemia in this age
group, we, similar to other investiga-
tors,35 did not identify an association
between growth and anemia. Measurement of children’s growth trajectories may have helped us identify
such an association, but unfortunately,
birth records were generally unavailable or unreliable. Our data did confirm previously reported associations
between lower hemoglobin levels and
male gender,36 findings that were possibly related to greater absolute longitudinal growth among boys.
The risk of iron-deficiency anemia may
thus depend on complex interactions
between dietary-iron content (type of
diet), iron bioavailability (duration of
breastfeeding and appropriate complementary feeding practices), increased iron use (growth velocity and
erythroid mass expansion), and inappropriate iron losses (infection and infestation).
Other biological factors associated
with childhood hemoglobin levels included serum folate level, presence of
inflammation, and hemoglobinopathy
status. We identified associations between hemoglobin and CRP levels only
when we included ferritin levels in the
regression equation; higher CRP levels
decreased the coefficient of the relationship between hemoglobin and ferritin levels, a result related to ferritin
being an acute-phase protein. Although vitamin A intake was often low
and was associated with hemoglobin
levels, serum RBP level was not associated with hemoglobin level, a finding
that may be related to a successful
government program to supplement
vitamin A.
The results of our study highlight important associations of wealth and food insecurity with anemia that, although previously reported,37 are independent of
other measured environmental and biological factors. This observation suggests that broader socioeconomic conditions directly influence hemoglobin
levels in children. Potential explanations
Downloaded from by guest on August 22, 2014
Child’s nutrition
• Complementary
food iron and
Pathway of factors associated with hemoglobin. Determinants of anemia, such as deficiencies in iron and folate, should be considered in the broader
context of family wealth, food insecurity, and associated nutritional intake and maternal nutritional status. Gender and hemoglobinopathies also contribute
to anemia.
include generalized bone marrow failure
because of malnutrition,38 deficiencies in
other micronutrients,39 exposure to biofuel smoke,40 and possibly other unexplained mechanisms associated with
lower socioeconomic status. We were
surprised that we found no association
between the highest level of food insecurity and hemoglobin level (Table 4), a result that may have been related to the
nonlinear performance of the Household
Food Insecurity Access Scale.19
Major consequences of climate change
are impairments of crop yield and agricultural productivity,41 conditions that
could increase food insecurity and
worsen childhood anemia. Furthermore,
the continuing global financial crisis
may threaten the health status of lowand middle-income countries42 and may
play a role in childhood anemia through
its effect on food insecurity. Thus, childhood anemia may worsen if the stressors listed above undermine socioeconomic advancement or worsen food
insecurity in India. Incorporation of strategies to support nutrition and address
socioeconomic conditions may help mitigate these phenomena.
Few studies conducted worldwide
have comprehensively examined the
etiology of anemia in children in the
infant-to-toddler age group. Although
to our knowledge there have been no
studies of anemia etiology in rural India, a study in urban slums of New
Delhi investigated 90 anemic children
and identified important contributions
from iron and vitamin B12 deficiency.43
A study of young Mexican children revealed that anemia attributable to
iron-deficiency anemia was less common than anemia from other causes.11
In Malawi, infectious diseases and vitamin B12 and folate deficiencies, but not
iron deficiency, were important factors associated with severe childhood
anemia.12 Results of studies in Thailand and the United States also indicated that iron deficiency was a
nondominant cause of pediatric anemia.44,45 In addition to confirming the
findings of the NFHS (which was a
large prevalence study), our data provide major insights into biological,
sociodemographic, and economic factors associated with anemia in rural
toddlers. The contrast between our
findings of predominant irondeficiency anemia and the findings reported in the published literature may
reflect differences in diet and socioeconomic patterns in this area and
also in study methods.
The findings of our study should be
considered with awareness of the following limitations. First, this is a crosssectional study, for which we report
association rather than causation.
Second, measurement of additional
laboratory variables, particularly levels of soluble transferrin receptor,
methylmalonic acid, and homocysteine, could have increased the detection of iron deficiency and functional
folate and vitamin B12 deficiencies.
Furthermore, we did not evaluate levels of lead46 or selenium,47, which have
been previously demonstrated to be
associated with anemia. These assays
were prohibited because the amount
of blood they require exceeded the
maximum phlebotomy volume acceptable to the community. Third, incomplete stool sampling in the field may
have resulted in failure to detect an
association between hookworm infestation and hemoglobin or ferritin levels. Finally, the 24-hour dietary-recall
method we used has limitations,21 and
may have led to overestimation of nutritional intake in young children compared with methods that use weight
measurement.48,49 However, this
method was the tool that could be
most feasibly administered within the
Downloaded from by guest on August 22, 2014
available time frame to assess nutritional intake in our field setting.
Despite these limitations, we have
identified a comprehensive set of factors associated with hemoglobin levels in rural Indian toddlers. The high
level of community participation, identification of anemia prevalence that reflects nationwide prevalence,2 and
socioeconomic similarities of the
selected districts of Karnataka with
other states in India suggest that the
results of this study may be generalizable to much of India and perhaps to
other resource-limited settings in Asia.
Iron-deficiency anemia, a leading risk
factor for burden of disease in developing countries,50 is associated with
impaired cognitive development7 and
potentially restricts economic development.51 Globally, policy makers have
deployed strategies against anemia
that include iron supplementation,
food fortification, and dietary diversification.52 Although the Indian anemiacontrol program recommends that
children younger than 5 years receive
iron and folic acid supplements,5,53 our
study results show that this approach
has not successfully controlled anemia prevalence. This apparent lack of
success may be related to suboptimal
program implementation,54 lack of
adherence,55 or other unidentified
causes. Thus, additional work is required to identify reasons for the gap
between policy and practice for anemia control in this setting.
The findings of our study support the
need for a broad public-health strategy for the control of anemia among
Indian children beyond delivering iron
supplementation alone. Measures that
address maternal anemia could have
functional and reproductive benefits
for mothers and, subsequently, children.56 The recent WHO recommendation to provide weekly iron and folic
acid to all women of reproductive age
could be expanded in rural India.57 Low
dietary-iron intake, particularly in
breastfeeding children, ideally should
be alleviated with a combined approach of iron supplementation, fortification of complementary foods, and
dietary education. These efforts must
be coupled with strategies to address
family poverty and food security, because both are independently associated with hemoglobin levels in
tention. Our findings suggest that
current public-health strategies such
as iron supplementation are necessary but not sufficient to reduce childhood anemia. Instead, combining iron
supplementation and food-fortification
programs with efforts to reduce maternal anemia, family poverty, and food
insecurity may yield optimal improvement of children’s hemoglobin levels.
Anemia, an important problem worldwide, is increasing among young children in India and requires urgent at-
This study was supported by grant
funding from the Allen Foundation,
Michigan (to Dr Shet) and funds from
the Fred P. Archer Charitable Trust,
Victoria, Australia, and the Melbourne
Research Scholarship, University of
Melbourne, Victoria, Australia (to Dr
We acknowledge the field team, led by
Mrs Varalaxmi Vijaykumar, which was
involved in data collection. We thank
Ms Shubha K. for assistance with data
entry. We are indebted to the community health workers and village Anganwadi workers who assisted with our
field work. We are grateful to Dr Julie
Simpson for guidance with the statistical analysis. We thank Professor Rob
Moodie for guidance in developing the
2001. Available at: Accessed April 11, 2010.
randomised controlled trials. Public Health
Nutr. 2005;8(2):117–132
4. Chatterjee P. Child malnutrition rises in India despite economic boom. Lancet. 2007;
8. Hurtado EK, Claussen AH, Scott KG. Early childhood anemia and mild or moderate mental retardation. Am J Clin Nutr. 1999;69(1):115–119
5. Kumar A. National nutritional anaemia control programme in India. Indian J Public
Health. 1999;43(1):3–5, 16
9. World Health Organization. Iron Deficiency
Anaemia: Assessment, Prevention, and
Control—A Guide for Programme Managers. Geneva, Switzerland: World Health
Organization; 2001. Available at: www.who.
index.html. Accesses April 12, 2010
1. Benoist B, McLean E, Egli I, Cogswell, eds.
Worldwide Prevalence of Anaemia
1993–2005. Geneva, Switzerland: World
Health Organization; 2008. Available at:
9789241596657㛭eng.pdf. Accessed April 11,
2. International Institute for Population Sciences and Macro International. National
Family Health Survey (NFHS-3), 2005–2006:
Key Findings. Mumbai, India: International
Institute for Population Sciences; 2007.
Available at:
pdf/SR128/SR128.pdf. Accessed April 11,
3. Registrar General and Census Commissioner, India. Census of India. New Delhi,
India: Office of the Registrar General, India;
PEDIATRICS Volume 126, Number 1, July 2010
6. International Institute for Population Sciences. The Second National Family Health
Survey (NFHS-2), 1998 –1999. Mumbai,
India: International Institute for Population
Sciences (IIPS) and ORC Macro; 2000. Available at:
Accessed April 11, 2010
7. Sachdev H, Gera T, Nestel P. Effect of iron
supplementation on mental and motor development in children: systematic review of
10. Thankachan P, Walczyk T, Muthayya S, et al.
Iron absorption in young Indian women: the
interaction of iron status with the influence
of tea and ascorbic acid. Am J Clin Nutr.
2008;87(4):881– 886
Downloaded from by guest on August 22, 2014
11. Duque X, Flores-Hernandez S, Flores-Huerta
S, et al. Prevalence of anemia and deficiency
of iron, folic acid, and zinc in children
younger than 2 years of age who use the
health services provided by the Mexican Social Security Institute. BMC Public Health.
12. Calis JC, Phiri KS, Faragher EB, et al. Severe
anemia in Malawian children. N Engl J Med.
2008;358(9):888 – 899
13. Schneider JM, Fujii ML, Lamp CL, et al. Anemia, iron deficiency, and iron deficiency
anemia in 12–36-mo-old children from lowincome families. Am J Clin Nutr. 2005;82(6):
1269 –1275
14. PHC Monthly Data: November 2007. Bangalore, India: Karuna Trust; 2008
15. Bhandari L, Kale S. Karnataka: Performance, Facts and Figures. Delhi, India: Indicus Analytics; 2007
16. Haahr M. Web site. Available at: Accessed April 12,
17. Pasricha SR, Vijaykumar V, Prashanth N, et
al. A community based field research
project investigating anaemia amongst
young children living in rural Karnataka,
India: a cross sectional study. BMC Public
Health. 2009;9(1):59
18. Swindale A, Bilinsky P. Development of a universally applicable household food insecurity measurement tool: process, current
status, and outstanding issues. J Nutr. 2006;
19. Coates J, Swindale A, Bilinsky P. Household
Food Insecurity Access Scale (HFIAS) for
Measurement of Household Food Access: Indicator Guide. Version 3. Washington, DC:
Food and Nutrition Technical Assistance
Project, Academy for Educational
Development; 2007. Available at: www. Accessed
April 12, 2010
20. Rutstein SO, Johnson K. The DHS Wealth Index. Calverton, MD: ORC Macro; 2004
21. Horst CH, Obermann-De Boer GL, Kromhout
D. Validity of the 24-hour recall method in
infancy: the Leiden Pre-school Children
Study. Int J Epidemiol. 1988;17(1):217–221
22. Gopalan C, Rama Sastri BV, Balasubramanian S. Updated by Narasinga Rao BS, Deosthale YG, Pant KC. Nutritive Value of Indian
Foods. Hyderabad, India: National Institute
of Nutrition, Indian Council of Medical
Research; 1989
23. World Health Organization. The WHO Child
Growth Standards. Geneva, Switzerland: Department of Nutrition, World Health
Organization; 2008. Available at: www.who.
int/childgrowth/standards/en. Accessed
April 12, 2010
Ash LR, Orihel TC, Savioli L, Abu Sin M, Montresor A, Renganathan E, eds. Training Manual
on Diagnosis of Intestinal Parasites: Tutor’s
Guide. Geneva, Switzerland: World Health
Organization; 2004. Available at:
en/trainingmanual㛭sip98-2.pdf. Accessed
April 12, 2010
Pulliam PN, Attia MW, Cronan KM. C-reactive
protein in febrile children 1 to 36 months of
age with clinically undetectable serious
bacterial infection. Pediatrics. 2001;108(6):
Bain BJ. Haemoglobinopathy Diagnosis. 2nd
ed. Oxford, United Kingdom: Blackwell
Publishing; 2006
de Pee S, Dary O. Biochemical indicators of
vitamin A deficiency: serum retinol and serum retinol binding protein. J Nutr. 2002;
132(9 suppl):2895S–2901S
Gamble MV, Ramakrishnan R, Palafox NA, et
al. Retinol binding protein as a surrogate
measure for serum retinol: studies in vitamin A-deficient children from the Republic
of the Marshall Islands. Am J Clin Nutr. 2001;
73(3):594 – 601
Kirkwood BR, Sterne JA. Essential Medical
Statistics. 2nd ed. Oxford, United Kingdom:
Blackwell Science; 2003
Bain B. Blood Cells: A Practical Guide. 4th ed.
Oxford, United Kingdom: Blackwell; 2006
Wharton BA. Iron deficiency in children: detection and prevention. Br J Haematol. 1999;
106(2):270 –280
Kumar A, Rai AK, Basu S, et al. Cord blood
and breast milk iron status in maternal anemia. Pediatrics. 2008;121(3). Available at:
Sharpe LM, Peacock WC, Cooke R, et al. The
effect of phytate and other food factors on
iron absorption. J Nutr. 1950;41(3):433– 446
Farley MA, Smith PD, Mahoney AW, et al.
Adult dietary characteristics affecting
iron intake: a comparison based on iron
density. J Am Diet Assoc. 1987;87(2):
184 –189
Chen K, Zhang X, Li TY, et al. Co-assessment
of iron, vitamin A and growth status to investigate anemia in preschool children in
suburb Chongqing, China. World J Pediatr.
Domellöf M, Lonnerdal B, Dewey KG, et al.
Gender differences in iron status during
infancy. Pediatrics. 2002;110(3):545–552
Skalicky A, Meyers AF, Adams WG, et al. Child
food insecurity and iron deficiency anemia
in low-income infants and toddlers in the
United States. Matern Child Health J. 2006;
Smith RR, Spivak JL. Marrow cell necrosis in
anorexia nervosa and involuntary starvation. Br J Haematol. 1985;60(3):525–530
Russell RM, Suter PM. Vitamin and trace
mineral deficiency and excess. In: Fauci AS,
Braunwald E, Kasper DL, et al, eds. Harrison’s Principles of Internal Medicine. 17th
ed. New York, NY: McGraw-Hill; 2008:
441– 450
Mishra V, Retherford RD. Does biofuel
smoke contribute to anaemia and stunting
in early childhood? Int J Epidemiol. 2007;
McMichael AJ, Powles JW, Butler CD, et al.
Food, livestock production, energy, climate
change, and health. Lancet. 2007;370(9594):
Horton R. The global financial crisis: an
acute threat to health. Lancet. 2009;
Gomber S, Kumar S, Rusia U, et al. Prevalence and etiology of nutritional anaemias
in early childhood in an urban slum. Indian
J Med Res. 1998;107:269 –273
Thurlow R, Winichagoon P, Green T, et al.
Only a small proportion of anemia in northeast Thai schoolchildren is associated with
iron deficiency. Am J Clin Nutr. 2005;(82):
380 –387
White KC. Anemia is a poor predictor of iron
deficiency among toddlers in the United
States: for heme the bell tolls. Pediatrics.
Jain NB, Laden F, Guller U, et al. Relation
between blood lead levels and childhood
anemia in India. Am J Epidemiol. 2005;
161(10):968 –973
Nhien NV, Khan NC, Yabutani T, et al. Relationship of low serum selenium to anemia
among primary school children living in rural Vietnam. J Nutr Sci Vitaminol (Tokyo).
2008;54(6):454 – 459
Fisher JO, Butte NF, Mendoza PM, et al. Overestimation of infant and toddler energy intake by 24-h recall compared with weighed
food records. Am J Clin Nutr. 2008;88(2):
407– 415
Olinto MT, Victora CG, Barros FC, et al.
Twenty-four-hour recall overestimates the
dietary intake of malnourished children. J
Nutr. 1995;125(4):880 – 884
Ezzati M, Lopez AD, Rodgers A, et al. Selected
major risk factors and global and regional
burden of disease. Lancet. 2002;360(9343):
Hunt JM. Reversing productivity losses
from iron deficiency: the economic case. J
Nutr. 2002;132(4 suppl):794S– 801S
Downloaded from by guest on August 22, 2014
52. Zimmermann MB, Hurrell RF. Nutritional
iron deficiency. Lancet. 2007;370(9586):
53. Saxena S. Review of the Policy Regarding
Micronutrients: Iron Folic Acid (IFA). New
Delhi, India: Ministry of Health and Family
Welfare, Government of India; 2007
54. Vijayaraghavan K, Brahmam GN, Nair KM, et
PEDIATRICS Volume 126, Number 1, July 2010
al. Evaluation of national nutritional anemia
prophylaxis programme. Indian J Pediatr.
55. Galloway R, McGuire J. Determinants of
compliance with iron supplementation:
supplies, side effects, or psychology? Soc
Sci Med. 1994;39(3):381–390
56. Allen LH. Anemia and iron deficiency: effects
on pregnancy outcome. Am J Clin Nutr.
2000;71(5 suppl):1280S–1284S
57. Vir SC, Singh N, Nigam AK, et al. Weekly iron
and folic acid supplementation with counseling reduces anemia in adolescent girls: a
large-scale effectiveness study in Uttar
Pradesh, India. Food Nutr Bull. 2008;29(3):
186 –194
Downloaded from by guest on August 22, 2014
Determinants of Anemia Among Young Children in Rural India
Sant-Rayn Pasricha, James Black, Sumithra Muthayya, Anita Shet, Vijay Bhat,
Savitha Nagaraj, N. S. Prashanth, H. Sudarshan, Beverley-Ann Biggs and Arun S.
Pediatrics 2010;126;e140; originally published online June 14, 2010;
DOI: 10.1542/peds.2009-3108
Updated Information &
including high resolution figures, can be found at:
This article cites 37 articles, 17 of which can be accessed free
This article has been cited by 4 HighWire-hosted articles:
Subspecialty Collections
This article, along with others on similar topics, appears in
the following collection(s):
Hospital Medicine
Patient Education/Patient Safety/Public Education
Permissions & Licensing
Information about reproducing this article in parts (figures,
tables) or in its entirety can be found online at:
Information about ordering reprints can be found online:
PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly
publication, it has been published continuously since 1948. PEDIATRICS is owned, published,
and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk
Grove Village, Illinois, 60007. Copyright © 2010 by the American Academy of Pediatrics. All
rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.
Downloaded from by guest on August 22, 2014