Received: 23 January 2023 | Revised: 6 December 2023 | Accepted: 9 December 2023

DOI: 10.1111/mcn.13619

ORIGINAL ARTICLE

Correlates of early child development among children with

stunting: A cross‐sectional study in Uganda

Joseph Mbabazi1,2 | Hannah Pesu1 | Rolland Mutumba1,2 | Kieran Bromley3 |

Christian Ritz4 | Suzanne Filteau5 | André Briend1,6 | Ezekiel Mupere2 |
Benedikte Grenov1 | Henrik Friis1 | Mette F. Olsen1,7
1
Department of Nutrition, Exercise & Sports, University of Copenhagen, Copenhagen, Denmark
2
Department of Paediatrics and Child Health, Makerere University, Kampala, Uganda
3
School of Medicine, Keele University, Keele, UK
4
The National Institute of Public Health, Southern University of Denmark, Copenhagen, Denmark
5
Faculty of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, London, UK
6
Tampere Center for Child Health Research, University of Tampere, Tampere, Finland
7
Department of Infectious Diseases, Rigshospitalet, Copenhagen, Denmark

Correspondence
Joseph Mbabazi, Department of Paediatrics Abstract
and Child Health, School of Medicine,
Many children in low‐ and middle‐income countries are not attaining their
Makerere University College of Health
Sciences, P.O Box 7062 Kampala, Uganda. developmental potential. Stunting is associated with poor child development, but
Email: mjosef2000@gmail.com
it is not known which correlates of stunting are impairing child development. We
Funding information explored potential socioeconomic, nutritional, clinical, and household correlates of
Arla Food for Health; Augustinus Fonden; early child development among 12–59‐month‐old children with stunting in a cross‐
Danish Dairy Research Foundation; A. P. Møller
Fonden til Lægevidenskabens Fremme; Læge sectional study in Uganda. Development was assessed using the Malawi Develop-
Sophus Carl Emil Friis og hustru Olga Doris ment Assessment Tool (MDAT) across four domains of gross and fine motor,
Friis' Legat
language, and social skills. Linear regression analysis was used to assess correlates of
development in the four domains and total MDAT score. Of 750 children included,
the median [interquartile range] age was 30 [23−41] months, 55% of the children
resided in rural settings with 21% from female‐headed households and 47% of
mothers had no schooling. The mean ± standard deviation height‐for‐age z‐score
(HAZ) was −3.02 ± 0.74, 40% of the children had a positive malaria test and 65%
were anaemic (haemoglobin < 110 g/L). One‐third had children's books at home,
majority (96%) used household objects to play with and most of them (70%) used
toys as pretence items like those to mimic cooking. After age, sex, and site
adjustments, HAZ (0.24, 95% confidence interval [CI]: 0.14−0.33) and head
circumference (0.07, 95% CI: 0.02−0.12) were positive correlates of total MDAT
score, whereas weight‐for‐height z‐score (WHZ) was not. Current breastfeeding was

Henrik Friis and Mette F. Olsen are contributed equally to this study.

This is an open access article under the terms of the Creative Commons Attribution‐NonCommercial License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited and is not used for commercial purposes.
© 2024 The Authors. Maternal & Child Nutrition published by John Wiley & Sons Ltd.

Matern Child Nutr. 2024;20:e13619. wileyonlinelibrary.com/journal/mcn | 1 of 12

https://doi.org/10.1111/mcn.13619
2 of 12 | MBABAZI ET AL.

associated with 0.41 (95% CI: 0.17−0.65) lower total MDAT score. Children from
households with a single income earner had 0.22 (95% CI: 0.06−0.37) lower total
MDAT score. Furthermore, severe food insecurity, inflammation and positive malaria
test were associated with lower scores for motor development. All family care
indicator subscales (FCIs) positively correlated with the total MDAT score and this
association was independent of household's socioeconomic status. In conclusion,
stunting degree, head circumference, number of household income earners and
stimulation by improved FCIs correlate with early child development among stunted
children. The negative association with prolonged breastfeeding is likely due to
reverse causality. Identified correlates may inform initiatives to support children
with stunting attain their development potential.

KEYWORDS
child development, correlation, Malawi Development Assessment Tool, stunting

1 | INTRODUCTION
Key messages
Around 250 million children (43%) under 5 years in low‐ and
middle‐income countries (LMICs) are at risk of not reaching their • Our findings reinforce the association between stunting

developmental potential (Black et al., 2017). Additionally, 22% of and early child development.

children under 5 years exhibit linear growth faltering (UNICEF & • The negative association with prolonged breastfeeding is

World Bank, 2023) which if were eliminated would possibly increase likely due to reverse causality as a result of mothers nursing

global annual income by ~$176.8 (Fink et al., 2016). The association their smaller, thinner, and more sickly children longer.

between stunting and poor child development is well documented • The strong correlation between all family care indicators

(Perkins et al., 2017). Stunted children tend to suffer short‐ and long‐ with development may express the benefit of improved

term consequences including increased morbidity and mortality, poor home stimulation such as provision of reading materials

cognitive, motor, and language development, and increased risk of and varied play materials on early child development

noncommunicable diseases later in life (Koshy et al., 2022; even among already stunted children in poor settings.

WHO, 2017a, 2017b).

Child development entails the physiological, psychological and
emotional changes that occur in humans from conception to the end
of adolescence as they progress from dependency to increasing
autonomy (Choo et al., 2019). Africa has a high burden of risk factors 2 | METHODS
for poor child development including intrauterine growth restriction,
stunting, iodine‐deficiency, iron‐deficiency anaemia, positive malaria 2.1 | Study design and participants
test, lead exposure, HIV, inadequate cognitive stimulation, maternal
depression and anaemia during pregnancy (Donald et al., 2019; Ford This was a cross‐sectional study using baseline data collected
& Stein, 2016; WHO, 2017a, 2017b). On the contrary, maternal between February and September 2020 as part of the MAGNUS
education, higher birth weight and socioeconomic status, and trial (Hannah Pesu et al., 2021) (ISRCTN13093195). As previously
breastfeeding have been identified as protective of child develop- described (Hannah Pesu et al., 2021; Mbabazi et al., 2023), the study
ment (Donald et al., 2019; Hadi et al., 2021). Although these have was conducted at two public health centres in Walukuba division and
mainly been studied among nonstunted children. Buwenge town council in Jinja district of East Central Uganda
However, it is unlikely stunting per se, but rather the (Busoga sub region). The current cross‐sectional study nested within
determinants and other correlates of stunting that impairs child the RCT uses the maximum available sample at baseline of 750
development (Leroy & Frongillo, 2019). Therefore, we explored the children. Participants were identified by village health teams in
socioeconomic, anthropometric, clinical and household correlates of nearby villages. Children with height‐for‐age z‐score (HAZ) <−2 and
early child development among children with stunting in an LMIC weight‐for‐height z‐score (WHZ) >−3, were referred to a study clinic
setting in Uganda. for assessment of eligibility. Those found with severe acute
MBABAZI ET AL. | 3 of 12

malnutrition (SAM) that is WHZ < −3, MUAC < 11.5, or bipedal pitting and Toddler Development (Bayley, 2006) as previously used to support
oedema were referred for treatment. The village‐level screening was MDAT assessment (Olsen et al., 2020; van den Heuvel et al., 2017). It
halted between 26 March to 14 June 2020 due to the global corona was assessed how happy, engaged, cooperative and anxious, children
pandemic, a period during which specific infection prevention and appeared during the assessment, and how much support their caregiver
control study operating procedure was established in compliance provided without necessarily influencing the test.
with the national and ethical committee regulations. The child's developmental stimulation at home was assessed
At the study clinic, children were included if they met the following using an African validated family care indicators (FCI) structured
eligibility criteria: lived within the catchment area, their caregivers were interview (Hamadani et al., 2010). The caregiver was asked about the
willing to return for follow‐up visits and agreed to home visits. Children level of stimulating home environment in four subscales including
with SAM, medical complications requiring hospitalisation, a history of availability of reading materials at home and their number, source of
peanut or milk allergies, overt disability impeding ability to eat or play materials, variety of play materials, and engagement with an
impeding measurement of length/height were excluded. Children were older family member (≥15 years) in various interactive activities over
also excluded if they participated in another study or if their family the past 3 days. These activities included reading books, telling
planned to move away from the catchment area. One child was stories, singing songs, taking the child out, playing with the child, and
recruited per household and in case of multiple stunted children, one counting or drawing with the child.
was randomly selected aided by phone app (True random generator).
Nonetheless, the caregiver was asked to take all the children to the
study clinic on their appointment day just in case the sampled child did 2.3 | Sociodemographic, socioeconomic
not meet inclusion criteria. Any siblings or twin under 5 years received and dietary intake data collection
similar supplements irrespective of their stunting status (unless if SAM)
but were not included in the study. A questionnaire was used to collect data on sociodemographic and
socioeconomic information. Breastfeeding status was assessed by
asking the mother if the child was still breastfeeding at the time of
2.2 | Assessment of child development and data collection. Dietary intake was based on 24‐h recall and dietary
household stimulation diversity was calculated based on the WHO global nutrition
monitoring framework operational guidance that recommends
Child development was assessed using the Malawi Development consumption of at least five out of eight food groups including
Assessment Tool (MDAT) version 6 (Gladstone et al., 2010) translated breastmilk in past 24 h (WHO, 2017a, 2017b). Minimum dietary
to both Lusoga and Luganda. The tool has been adapted and diversity has been documented to be associated with socioeconomic
validated for use in LMICs including Uganda and was developed in an status in low‐income settings (Scarpa et al., 2022). Food security was
African setting. It focuses on four domains including gross motor, fine calculated using the USAID household food insecurity access scale
motor, language, and social skills with 39, 42, 40 and 36 milestones in (Coates et al., 2007). All caregivers received nutrition counselling
each domain, respectively. The MDAT is primarily an observation‐ using the national guidelines on infant and young child feeding
based tool with standardised items assessed by a trained research (MOH, 2009).
assistant referred to as a child development officer (CDO). During
assessments, most of the items were observed while some, mainly in
the social domain, were assessed based on caregiver report. Normal 2.4 | Anthropometric measures
age‐specific reference values for each domain were used as a starting
point while testing each child. The CDO first performed a forward All anthropometric measurements were repeated in triplicate and the
test until the child failed six consecutive items thereby marking the median used. Weight was taken using an electronic scale with double
rest of the items above as failed. If the child passed six consecutive weighing function (Seca 876). Length (measured in children < 24
items in the forward test, all items below were marked passed; months) and height (measured in children 24 months and older) was
otherwise, a backward test was performed until six consecutive items taken using a wooden ShorrBoard® (Weigh and Measure) ensuring 5‐
were passed. To minimise distraction, child development assessments points of contact with repositioning between measurements. Mid‐
took place in a separate tent located away from the rest of the upper arm circumference (MUAC) and head circumference were
stations. After every 20 child development assessments per CDO, a measured using a nonelastic MUAC tape (UNICEF SD) and head
quality check was performed by another CDO concurrently and circumference tape (Seca 212).
results were compared. In case of any discrepancy, consensus was
arrived at in consultation with the respective standard operating
procedures and views from other CDOs. 2.5 | Blood sampling and analysis
The child's participation during the MDAT assessment was
observed by the CDO. This was evaluated based on an adapted version Venous blood was drawn from each child, transported to the field
of the Behaviour Observation Inventory from the Bayley Scale of Infant laboratory, processed, and temporarily stored at −20°C before being
4 of 12 | MBABAZI ET AL.

transported to Kampala for storage at −80°C. Processed samples 2.7 | Ethical statement
were later transferred to Denmark and Germany on dry ice for
analysis of the micronutrient biomarkers and acute phase proteins as The study was approved by the School of Medicine Research and
elaborated elsewhere including justification for cut‐offs (Mutumba Ethics Committee of Makerere University, Kampala and the
et al., 2023). Before processing, whole blood was used to diagnose Uganda National Council of Science and Technology. A consulta-
malaria (rapid diagnostic test RDT, SD bioline malaria Ag Pf, Abbott) tive approval was obtained from the Danish National Committee
and measure haemoglobin concentration (Hb201+, HemoCue). on Biomedical Research Ethics. The study was conducted in
accordance with the principles of the Helsinki Declaration (World
Medical Association, 2014) and followed local guidelines for
2.6 | Data management and statistical analyses human research. All study staff undertook a course in Good
Clinical Practice and Human Subject Protection. Oral and written
Data was collected using paper case report forms and double entered information was provided in Lusoga, Luganda or English. Before
in EpiData before submission to a secure server periodically. Statistical caregivers gave written informed consent, their understanding
analysis was done using Stata SE 14 (StataCorp LP). Descriptive of the information was checked by a different staff member, using
statistics are presented as mean ± SD, median [interquartile range] and a questionnaire.
frequency, n (%). We generated internal MDAT developmental age‐
adjusted z‐scores using the generalised additive model for location
scale and shape (GAMLSS) (Stasinopoulos et al., 2022) in R software. 3 | RESULTS
This model enables fitting regression models where the distribution of
the outcome does not fit a typical distribution and allows for fine‐ In this study, 750 children with stunting were enroled. The median
tuning of the location, scale, and shape parameters in its construction, [IQR] age was 30 [23−41] months and slightly over half of them
to better adapt the model to the shape of the data. resided in rural settings (Table 1). Three‐quarters of the households
Internal MDAT scores were generated using an item response theory spent more than half of their income on food, nearly half of the
(IRT) analysis through the use of unidimensional 2‐paramter‐logistic (2PL) children had mothers who had not been to school, and one fifth lived
models based upon responses to the MDAT. Thereafter, GAMLSS were in female‐headed households. Of the 95 (12.7%) currently breastfed
utilised to then generate an age‐contingent measure of ability based on children, 86 (38.7%) were among 222 children between 12 and 23
the development scores from the IRT analysis, thus removing the impact months and the remaining 9 (3.5%) among 256 children between
of age on development, and converting the outcome to a z‐score. 24 and 35 months. Almost half of the children were severely stunted
Linear regression analysis was used to assess correlates of gross with a mean ± SD HAZ of −3.02 ± 0.74, more than One‐third had
motor, fine motor, language development, social skills and the total positive malaria test and about two thirds were anaemic.
MDAT z‐scores. Potential correlates included sociodemographic The mean ± SD total MDAT score was −0.20 ± 1.00, gross motor:
characteristics (age, sex, current breastfeeding, urban residence and −0.19 ± 1.02, fine motor: −0.13 ± 1.04, language: −0.15 ± 1.03, and
household size), socioeconomic factors (income earners, food social skills: −0.13 ± 1.03. During the assessments, 654 (87%) of the
expenditure, food insecurity and dietary diversity) and anthropome- children were assessed as cooperative, 544 (73%) were happy, 714
try (head circumference, HAZ, WHZ and WAZ). We also explored (95%) very engaged, and 683 (91%) unfearful and without signs of
micronutrient status and other clinical factors for their correlation anxiety. Additionally, 593 (79%) of the caregivers were assessed as
with child development. These included haemoglobin (Hb), anaemia being supportive to their children during the MDAT assessment.
(Hb < 110 g/L), ferritin, iron deficiency anaemia (IDA) (Hb < 110 g/L One‐third had children's books at home. Nearly all used
and serum ferritin < 12 µg/L), serum soluble transferrin receptors household objects as playing materials, about half used home‐made
(>8.3 mg/L), inflammatory markers serum C‐reactive protein (CRP, toys and one‐third had purchased toys (Table 2). Out of the seven
>10 mg/L) and α−1‐acid glycoprotein (AGP, ≥1 g/L), plasma cobala- varieties of play materials assessed, a large proportion (70%) used
min (<148 pmol/l), plasma methylmalonic acid (MMA, >0.75 µmol/L) imitational materials such as things for pretending to cook, while only
and malaria (positive RDT). The cutoffs used are as previously few (3%) used toys for learning shapes or colours. Across the six
described (Mutumba et al., 2023) and serum ferritin was corrected for interactive activities that children might engage in with older family
inflammation as described elsewhere (Cichon et al., 2018). members, ‘reading books’ and ‘counting or drawing’ were uncommon
Our analysis included adjustment for age and sex as fixed effects with only 20% and 29% engaged in these two, respectively
and sites as random effects to obtain adjusted measures at 95% (Supporting Information S1: Figure 1). It was mainly mothers who
confidence interval (CI) and p < 0.05 significance level. FCIs, as proxies engaged with the children (49%) followed by other relatives (36%)
for household stimulation, were assessed for their correlation with child while the fathers were the least engaged, with only 12% found to
development using linear regression analysis. Furthermore, we exam- engage with their children in any interactive activity at home.
ined the correlation between the socioeconomic factors with the four Notably, 4% of the children had no engagement whatsoever with any
respective FCI subscales of (a) children's books at home, (b) sources of older family member during the past 3 days (Supporting Information
play materials, (c) variety of play materials, and (d) family interaction. S1: Figure 1).
MBABAZI ET AL. | 5 of 12

TABLE 1 Baseline characteristics of 750 children with stunting. TABLE 2 Family care indicators of 750 children with stunting.

Sociodemographic data Children's books at home

Age (months) 30 [23−41] None 66.9% (501)

Male sex 54.9% (412) Any books 33.2% (249)

Rural residence 55.2% (415) Sources of play materials

Household size 5 [4−7] Home‐made toys 55.3% (415)

Socioeconomic status and diet Purchased toys 36.7% (275)

Single income earner 70.2% (526) Household objects 96.3% (722)

a
Female‐headed households 21.3% (157) Summary score 1.9 ± 0.7

No maternal schooling 47.4% (338) Varieties of play materials

Income spent on food >50% 68.4% (513) Things that can make or play music 10.3% (77)

Severe household food insecurity (Access) 63.7% (478) Drawing or writing objects 61.5% (461)

Access to agricultural land 41.5% (311) Children's picture books 25.6% (192)
a
Diverse diet 26.3% (196) Things meant for stacking, construction, or building 59.1% (443)

Currently breastfed 12.7% (95) Things for moving around (e.g., bicycles) 41.9% (314)

Anthropometric data Toys for learning shapes or colours 2.9% (22)

Mid‐upper arm circumference (cm) 14.4 ± 1.2 Things for pretending (e.g., to pretend to cook) 70.3% (527)
b
Head circumference (cm) 47.2 ± 1.8 Summary score 2.7 ± 1.6

Height‐for‐age (z‐score) −3.02 ± 0.74 Family interaction

Weight‐for‐height (z‐score) −0.36 ± 0.99 Read books or looked at picture books 19.7% (148)

Weight‐for‐age (z‐score) −1.93 ± 0.85 Tell stories 36.0% (270)

Micronutrient and clinical data Sing songs 73.7% (553)

Positive malaria rapid diagnostic test 39.6% (292) Take your child outside the home 56.5% (424)

Haemoglobin < 110 g/L 64.5% (479) Played with the child 81.2% (609)

Serum C‐reactive protein > 10 mg/L 22.0% (163) Counted or drew things with the child 28.5% (214)

Serum α−1‐acid glycoprotein ≥ 1 g/L 63.6% (471) Summary scorec 3.0 ± 1.4

Serum ferritinb < 12 µg/L 42.9% (318) Note: For sources of play materials, varieties of play materials, and family
interaction, numbers exceed 100 as it was possible to choose more than
Plasma cobalamin <148 pmol/L 3.5% (25)
one option. Data reported as % (n) or mean ± SD.
Plasma methylmalonic acid >0.75 µmol/L 15.8% (116) a
Range from 0 (no source of play material) to 3 (having all source of play
materials).
Note: Data reported as mean ± SD, median [IQR] and % (n). b
a Range from 0 (no varieties play material) to 7 (having all varieties of play
Based on minimum dietary diversity score, eating from 5+ of 8 food
materials).
groups including breastmilk in the past 24 h. c
b Range from 0 (no family interaction) to 6 (family interaction in all
Corrected for inflammation (Cichon et al., 2018).
activities).

3.1 | Anthropometric, sociodemographic and CI: −0.65 to −0.17) than children who had been weaned, with
socioeconomic correlates negative associations in all the domains. The association between
breastfeeding and MDAT scores was not different between those
After adjustments for age, sex and site, HAZ was a positive below and above 24 months (interaction, p > 0.113). For the total
correlate of total MDAT score, whereas WHZ was not (Table 3a). MDAT score, the associations were −0.32 (95% CI: −0.59 to −0.05)
For HAZ, the total MDAT score was 0.24 (95% CI: 0.14−0.33) and −0.73 (95% CI: −1.39 to −0.06) respectively (interaction
higher for each 1‐unit higher HAZ and reflected associations in all p = 0.272). Conversely, compared to the nonbreastfed, currently
the four MDAT domains. Head circumference was also a correlate breastfed children generally had poorer anthropometric indices, with
of total MDAT score (0.07, 95% CI: 0.02−0.12), driven by bigger proportion of anaemic and cobalamin deficient cases (Table S1).
associations with gross motor and language scores. Children who Additionally, children from households with a single income earner had
were currently breastfed had lower total MDAT score (−0.41, 95% 0.22 (95% CI: 0.06−0.37) lower total MDAT score. This was driven by
 6 of 12
|

T A B L E 3a Anthropometric, sociodemographic and socioeconomic correlates of child development among 750 children with stunting.
Gross motor Fine motor Language Social skills Total score
Characteristic β (95% CI) P β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p

Anthropometric data

Height‐for‐age z‐score 0.28 (0.18−0.37) <0.001 0.16 (0.06−0.26) 0.002 0.15 (0.05−0.25) 0.003 0.12 (0.02−0.22) 0.02 0.24 (0.14−0.33) <0.001

Weight‐for‐height z‐score 0.05 (−0.02 to 0.13) 0.16 0.01 (−0.06 to 0.09) 0.71 0.03 (−0.04 to 0.11) 0.40 0.02 (−0.06 to 0.09) 0.68 0.04 (−0.03 to 0.11) 0.29

Head circumference (cm) 0.07 (0.02−0.13) 0.008 0.04 (−0.01 to 0.10) 0.13 0.07 (0.01−0.12) 0.01 0.03 (−0.03 to 0.08) 0.31 0.07 (0.02−0.12) 0.01

Sociodemographic status

Age (years) 0.08 (0.01−0.16) 0.03 0.06 (−0.02 to 0.14) 0.15 0.06 (−0.01 to 0.14) 0.11 0.01 (−0.07 to 0.09) 0.80 0.06 (−0.01 to 0.14) 0.095

Girl sex −0.14 (−0.29 to 0.003) 0.05 −0.05 (−0.20 to 0.10) 0.53 −0.08 (−0.23 to 0.07) 0.30 0.22 (0.07−0.37) 0.004 −0.06 (−0.21 to 0.08) 0.38

Urban residence −0.08 (−0.32 to 0.16) 0.52 −0.12 (−0.37 to 0.18) 0.34 −0.09 (−0.34 to 0.15) 0.45 −0.05 (−0.30 to 0.19) 0.68 −0.11 (−0.35 to 0.13) 0.36

Household size > 5 people −0.09 (−0.24 to 0.06) 0.23 −0.04 (−0.20 to 0.11) 0.61 −0.06 (−0.21 to 0.10) 0.47 0.13 (−0.02 to 0.29) 0.08 −0.03 (−0.18 to 0.12) 0.70

Socioeconomic status

Single income earner −0.19 (−0.35 to −0.03) 0.02 −0.15 (−0.31 to 0.02) 0.08 −0.18 (−0.34 to −0.02) 0.03 −0.12 (−0.28 to 0.05) 0.17 −0.22 (−0.37 to −0.06) 0.008

Income spent on food (>50%) 0.08 (−0.10 to 0.25) 0.38 0.12 (−0.06 to 0.30) 0.18 −0.07 (−0.25 to 0.10) 0.42 −0.02 (−0.20 to 0.15) 0.81 0.06 (−0.11 to 0.24) 0.45

Food insecurity

Mildly −0.38 (−0.90 to 0.13) 0.15 −0.33 (−0.86 to 0.21) 0.23 0.07 (−0.46 to 0.59) 0.80 −0.17 (−0.69 to 0.36) 0.53 −0.20 (−0.71 to 0.31) 0.43

Moderately −0.25 (−0.62 to 0.13) 0.20 −0.08 (−0.47 to 0.31) 0.69 −0.14 (−0.53 to 0.24) 0.47 0.08 (−0.30 to 0.46) 0.68 −0.15 (−0.52 to 0.22) 0.44

Severely −0.47 (−0.83 to −0.11) 0.01 −0.18 (−0.55 to 0.19) 0.35 −0.26 (−0.63 to 0.11) 0.17 −0.02 (−0.39 to 0.34) 0.91 −0.30 (−0.65 to 0.06) 0.10

No diverse diet 0.11 (−0.17 to 0.39) 0.44 0.08 (−0.21 to 0.37) 0.60 −0.16 (−0.44 to 0.13) 0.28 −0.09 (−0.37 to 0.20) 0.54 0.01 (−0.27 to 0.29) 0.93

Female‐headed households −0.09 (−0.27 to 0.09) 0.35 −0.04 (−0.23 to 0.14) 0.65 −0.04 (−0.23 to 0.14) 0.64 0.07 (−0.11 to 0.25) 0.46 −0.07 (−0.25 to 0.11) 0.45

No maternal schooling −0.11 (−0.26 to 0.04) 0.15 −0.11 (−0.27 to 0.04) 0.15 −0.06 (−0.21 to 0.10) 0.47 0.05 (−0.10 to 0.20) 0.54 −0.13 (−0.28 to 0.01) 0.08

Currently breastfed −0.31 (−0.55 to −0.06) 0.01 −0.25 (−0.50 to 0.003) 0.05 −0.32 (−0.57 to −0.07) 0.01 −0.55 (−0.80 to −0.30) <0.001 −0.41 (−0.65 to −0.17) 0.001

Note: Data reported as regression coefficients (95% CI), adjusted for age, sex and study site. For categorical variables, the following references were used: girls versus boys, currently versus not currently
breastfed, urban versus rural residence; household size >5 versus ≤5; single versus more than one income earner, income spent on food < versus ≥50%, food insecurity versus food secure; no diverse versus
diverse diet.
MBABAZI
ET AL.
MBABAZI ET AL. | 7 of 12

negative associations with language, gross and fine motor although the associated. Children from households with a single income earner and
latter was only marginally significant. Finally, severe household food those that were breastfed had lower development scores. Similarly,
insecurity was associated with lower developmental scores in the we found that lower scores in some specific child development
gross motor domain (−0.47, 95% CI: −0.83 to −0.11). domains were associated with inflammation, cobalamin status, and
positive malaria test. Finally, we found all indicators of family care to
be associated with higher child development scores and that this was
3.2 | Micronutrient and clinical correlates independent of household's SES.
Our findings concur with recent literature that has reported
Neither Hb nor any of the micronutrient or inflammatory markers stunting to be associated with delayed development with odds ratios
were associated with total MDAT score (Table 3b). Elevated serum (ORs) ranging between 3.0 and 3.71 (Hikmahrachim & Ronoatmodjo,
AGP and positive malaria RDT were negative correlates of gross 2020; Oumer et al., 2022; Rosyidah et al., 2021). A meta‐analysis
motor scores. Whereas very low plasma cobalamin was insignificantly noted stunting increased the risk of abnormal development 3.7 times
associated with low development, elevated plasma MMA, a specific compared to nonstunted < 5 year‐old children (Rosyidah et al., 2021).
marker of cobalamin deficiency, was associated with lower fine Findings from other studies have confirmed both growth faltering
motor score. and head circumference to be related to cognitive outcomes at 2
years (Scharf et al., 2018) as well as in older children aged 6−12 years
(Mutapi et al., 2021; Poh et al., 2013), and among HIV‐exposed but
3.3 | Household correlates uninfected children (Sirajee et al., 2021).
The negative association between current breastfeeding and
All FCIs were associated positively with total MDAT score (Table 3c) child development is contrary to reports from previous studies,
For every 1 more children's book owned at home, there was an where breastfeeding has been associated with better development,
associated 0.23 (95% CI: 0.08−0.38) higher total MDAT score. especially motor skills, with effect sizes ranging from 0.5 to 10.9
Additionally, having more than 2 sources of play materials, more than between breastfed versus nonbreastfed (Bernardo et al., 2013; Horta
3 varieties of play materials and more than 3 interactions with family et al., 2015; Quinn et al., 2001; Sacker et al., 2006). However, the
members were associated with 0.18 (95% CI: 0.002−0.37), 0.40 (95% results of these studies are prone to confounding, and in adjusted
CI: 0.24−0.55) and 0.23 (95% CI: 0.07−0.38) higher total MDAT analyses, the effect sizes between breastfed versus nonbreastfed
scores, respectively. The >3 varied play materials and >3 family tend to be attenuated: 0.4–8.2 (Bernardo et al., 2013; Hernández‐
interactions had positive associations across all the four MDAT Luengo et al., 2022; Horta et al., 2007; Horta et al., 2015; Quinn
domains. Having children's books at home correlated with gross and et al., 2001; Sacker et al., 2006). In fact, Renee et al. who studied the
fine motor, and language skills (marginally) but did not correlate with extent to which confounding explained the association between
acquisition of social skills. On the contrary, having >2 sources of play breastfeeding duration and cognitive development up to 14 years of
materials correlated with language and social skills acquisition age, reported that adjusting for SES tended to halve the effect sizes,
(marginally) but did not correlate with gross and fine motor while further adjustment for maternal cognitive scores explained the
development. The observed associations were independent of the remaining association at 5 years of age, while there was still an effect
households' SES and thus were retained after further adjustments for at ages 7, 11 and 14 (Pereyra‐Elías et al., 2022). In line with this, a
maternal schooling and/or number of income earners (Supporting systematic review of 17 studies in Sub‐Saharan Africa concluded that
Information S1: Table 2). much of the effect of breastfeeding that is, exclusive or prolonged
Further exploration of the characteristics of stimulative breastfeeding on child development has to do with ‘who breastfeeds’
households, children from those households with a single income and that confounders are rarely adequately considered in analyses
earner and no maternal schooling were negatively associated with (Mohammed et al., 2022). In fact, prolonged breastfeeding in itself
all the four FCIs (Supporting Information S1: Table 3). In addition, has been associated with undernutrition in several studies (Briend &
urban residence was associated with having more sources of, and Bari, 1989; Syeda et al., 2021). One study reported the odds of
varied play materials. We did not observe any differences by sex severe development delay were higher among children breastfed in
and estimates suggested that older children received more third year (adjusted OR, AOR: 6.19, 95% CI: 3.31−11.56) compared
stimulation especially with regard to sources and variety of play to children breastfed in their second year of life (AOR: 2.84, 95% CI:
materials. 1.18−4.46) (Syeda et al., 2021). In another systematic review, all
studies identified poor growth and nutritional status in children
breastfed for >1 year (Lackey et al., 2021). The negative association
4 | DISC US SION between prolonged breastfeeding and nutritional status could be a
result of reverse causality. This has been demonstrated in studies
In our study among 1–5‐year‐old children with stunting, linear (Caulfield et al., 1996; Grummer‐Strawn, 1993; Marquis et al., 1997)
growth status and head circumference positively correlated with showing that still breastfed children were shorter and lighter than
child development scores, whereas ponderal growth status was not their nonbreastfed counterparts of the same age. This is in line with
 8 of 12
|

T A B L E 3b Micronutrient and clinical correlates of child development among 750 children with stunting.

Gross motor Fine motor Language Social skills Total score

Characteristics β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p

Haemoglobin (Hb, g/L) 0.005 (−0.001 to 0.01) 0.08 −0.00 (−0.01 to 0.004) 0.90 0.002 (−0.003 to 0.01) 0.42 −0.00 (−0.01 to 0.01) 0.98 0.002 (−0.002 to 0.01) 0.29

Anaemia (Hb <110 g/L) −0.06 (−0.22 to 0.09) 0.42 0.04 (−0.12 to 0.20) 0.61 −0.02 (−0.18 to 0.14) 0.83 0.07 (−0.09 to 0.22) 0.41 −0.02 (−0.17 to 0.13) 0.81
a
Serum ferritin (µg/L) −0.14 (−0.35 to 0.06) 0.18 −0.01 (−0.22 to 0.21) 0.94 −0.04 (−0.25 to 0.17) 0.72 0.08 (−0.13 to 0.29) 0.45 −0.09 (−0.29 to 0.12) 0.40

24+ Ref Ref Ref Ref Ref

12–<24 0.15 (−0.04 to 0.35) 0.13 0.06 (−0.14 to 0.27) 0.55 0.15 (−0.05 to 0.35) 0.14 0.06 (−0.14 to 0.26) 0.53 0.17 (−0.02 to 0.36) 0.08

<12 0.12 (−0.08 to 0.32) 0.26 0.01 (−0.20 to 0.21) 0.95 0.09 (−0.12 to 0.29) 0.41 0.002 (−0.20 to 0.21) 0.98 0.12 (−0.07 to 0.32) 0.22

Serum transferrin 0.10 (−0.06 to 0.25) 0.23 0.04 (−0.13 to 0.20) 0.66 −0.004 (−0.16 to 0.16) 0.96 0.10 (−0.06 to 0.26) 0.21 0.06 (−0.10 to 0.21) 0.46
receptor >8.3 mg/L

Serum AGP (g/L)

<0.8 Ref Ref Ref Ref Ref

0.8−1.2 −0.18 (−0.40 to 0.03) 0.09 −0.04 (−0.26 to 0.18) 0.74 0.18 (−0.04 to 0.40) 0.10 −0.12 (−0.33 to 0.10) 0.29 0.02 (−0.19 to 0.23) 0.82

1.2+ −0.29 (−0.49 to −0.09) 0.01 −0.06 (−0.27 to 0.15) 0.59 0.03 (−0.18 to 0.24) 0.80 −0.15 (−0.36 to 0.06) 0.16 −0.13 (−0.33 to 0.08) 0.22

Plasma cobalamin (pmol/L)

222+ Ref Ref Ref Ref Ref

148−<222 0.12 (−0.07 to 0.31) 0.21 0.10 (−0.10 to 0.30) 0.33 0.09 (−0.11 to 0.28) 0.37 0.32 (0.13−0.51) 0.001 0.14 (−0.04 to 0.33) 0.13

<148 0.03 (−0.38 to 0.43) 0.90 −0.03 (−0.45 to 0.39) 0.89 −0.04 (−0.45 to 0.37) 0.85 0.14 (−0.27 to 0.54) 0.51 −0.02 (−0.41 to 0.38) 0.93

Plasma MMA (µmol/L)

<0.45 Ref Ref Ref Ref Ref

0.45–0.75 −0.08 (−0.29 to 0.12) 0.42 −0.27 (−0.48 to −0.07) 0.01 −0.16 (−0.36 to 0.05) 0.14 0.02 (−0.18 to 0.22) 0.85 −0.18 (−0.38 to 0.02) 0.07

>0.75 −0.10 (−0.31 to 0.11) 0.36 −0.22 (−0.44 to 0.001) 0.05 −0.002 (−0.22 to 0.22) 0.98 0.14 (−0.07 to 0.36) 0.19 −0.10 (−0.31 to 0.11) 0.34

Malaria (RDT positive) −0.17 (−0.33 to −0.01) 0.03 0.02 (−0.14 to 0.14) 0.80 −0.13 (−0.29 to 0.04) 0.13 −0.02 (−0.18 to 0.14) 0.80 −0.13 (−0.28 to 0.03) 0.11

Note: Data reported as regression coefficients (95% CI), adjusted for age, sex, and study site.
Abbreviations: AGP, α‐1‐acid glycoprotein; MMA, methylmalonic acid; RDT, rapid diagnostic test.
a
Corrected for inflammation (Cichon et al., 2018).
MBABAZI
ET AL.
MBABAZI ET AL. | 9 of 12

our findings including by (Mutumba et al., 2023), given that on top of

<0.001
0.003

0.048

0.004
poorer anthropometric indices, breastfed children had more anaemia
p and cobalamin deficient cases compared to the nonbreastfed, all
suggesting that the poor development was not due to breastfeeding

0.18 (0.002−0.37)
0.23 (0.08−0.38)

0.40 (0.24−0.55)

0.23 (0.07−0.38)
per se but rather the prolonged breastfeeding was being used
Total score

as a countermeasure to other nutritional, clinical or physiological

β (95% CI)

disruptions.
Our findings that inflammation and a positive malaria test were
associated with lower gross motor scores concur with recent cohort
<0.001

studies among mother−child dyads and younger children < 2 years

0.001
0.12

0.10

(Heesemann et al., 2021; Olsen et al., 2020) that also found

p

elevated serum AGP and positive malaria tests to be associated with

0.13 (−0.03 to 0.28)

0.16 (−0.03 to 0.35)

lower developmental scores. It should be noted that malaria remains

0.30 (0.13−0.46)

0.28 (0.12−0.44)

one of the greatest contributors to morbidity and mortality in sub‐

Social skills

Saharan Africa, including Uganda where almost half of children had

β (95% CI)

a positive malaria RDT in the 2016 DHS (ICF, U. B. o. S. U. a., 2017).

Moreover, a study by Milner et al. further revealed that the
association between malaria infection and early childhood develop-
<0.001

ment was actually mediated by anaemia (Milner et al., 2020). We did

0.06

0.04

0.02

not find Hb or anaemia to be associated with child development

p

scores. Previous studies have reported inconsistent results; studies

0.15 (−0.004 to 0.31)

in India (Heesemann et al., 2021) and Burkina Faso (Prado

0.20 (0.01−0.39)

0.37 (0.21−0.54)

0.19 (0.03−0.35)

et al., 2017), respectively, found higher Hb to be associated with

improved child development and higher test scores, while a
Family care indicators as correlates of child development among 750 children with stunting.

β (95% CI)
Language

Ugandan study found lower Hb levels to be associated with reduced

psychomotor scores at 15 months of age and no evidence that
anaemia or IDA were associated with cognitive or motor scores at
5 years (Nampijja et al., 2022).
0.003
0.02

0.11

0.09

We found all FCI subscales to correlate with higher total MDAT

p

Note: Data reported as regression coefficients (95% CI), adjusted for age, sex and study site.

score. Our results concur with findings from the iLiNS trial where
0.16 (−0.04 to 0.35)

0.14 (−0.02 to 0.30)

variety of play materials positively correlated with both language

0.19 (0.03−0.35)

0.25 (0.09−0.42)

and motor development (Prado et al., 2017). Indicators of family

Fine motor

care are crucial for ascertaining whether families provide their

β (95% CI)

children with an enabling environment that would in turn lead to

positive developmental outcomes. Results from our exploratory
analysis showed that households with a single income earner and no
<0.001

maternal schooling correlated with poor FCIs across all four

0.001

0.55

0.01

subscales. Poor home stimulation is a key factor affecting develop-

p

ment of children living in impoverished areas like LMICs (Black

0.06 (−0.13 to 0.24)

et al., 2017). With chronic undernutrition, this becomes double‐

0.27 (0.11−0.42)

0.32 (0.15−0.48)

0.22 (0.06−0.38)

jeopardy as such children are at high risk of developmental delay.

Gross motor

Despite challenges in ascertaining as to whether it's poverty or lack

β (95% CI)

of stimulation affecting development, understanding the stunted

children's home environment is key for planning and selecting
appropriate mitigation measures.
Sources of play materials > 2
Any children's book at home

Variety of play materials > 3

Our findings on children's interaction with older family members

concur with some previous studies alluding that compared to fathers,
Family interaction > 3

mothers interact more with the children (McBride & Mills, 1993).
Children experience rapid growth and development during early
years as they learn new things vital for their future success.
T A B L E 3c

Prospective studies have shown that family environment, and play

are associated with literacy skills (Nyah, 2021). Moreover, mothers
have been shown to participate in childrearing activities at higher
10 of 12 | MBABAZI ET AL.

rates than fathers and this pattern holds for both dual‐earner and ACKNOWL EDGEM ENTS
single‐earner families (McBride & Mills, 1993). Notably, fewer Henrik Friis has received research grants from ARLA Food for Health.
children with stunting in our study had access to books (33%), Henrik Friis and Benedikte Grenov have received research grants from
moreover with low motor scores which could partly explain the the Danish Dairy Research Foundation. Henrik Friis, Suzanne Filteau and
limited reading, and counting/drawing interaction as observed being André Briend have had research collaboration with Nutriset, a producer
the activity with the highest proportion of children not engaged by of LNS. Other authors declare no financial relationships with any
their older family members. This is alluded to in a study among organisations that might have an interest in the submitted work in the
Jamaican children on behaviour and development of stunted and previous 5 years, and declare no any other relationships or activities that
nonstunted children in which the former were more apathetic, with could appear to have influenced the submitted work.
less enthusiasm and varied exploration, less happy and more fussy
than the latter (Gardner et al., 1999). CONFLIC T OF INTEREST STATEM ENT
Among the key strengths of our study include the large sample The authors declare no conflict of interest.
size of stunted children, the wide range of exposures explored and
the use of an African developed tool validated in Uganda. However, DATA AVAILABILITY STATEMENT
the cross‐sectional nature of our study limits the ability to establish The Ugandan act on Data Protection and Privacy and the European
causal relationships and having studied only stunted children limits act on General Data Protection Regulation do not allow for personal
comparability with studies among nonstunted children. We relied on data to be made available to other researchers without prior written
caregiver reports for such information like on FCIs, which may be approval from relevant institutions and authorities. The Data
subject to recall and information bias. Protection Officer of the University of Copenhagen can be contacted
about data inquiries at dpo@adm.ku.dk.

5 | C ONC LUS I ON A N D ORC I D

R E C O MM E N D A T IO N S Joseph Mbabazi http://orcid.org/0000-0001-7602-8008
Rolland Mutumba http://orcid.org/0000-0002-1311-9946
We found that among children with stunting, HAZ was positively Suzanne Filteau http://orcid.org/0000-0002-1119-6825
associated with all child development domains. FCIs were consis- Benedikte Grenov http://orcid.org/0000-0003-0259-7851
tently correlated with better scores, independent of socioeconomic
status. If these associations reflect cause‐effect relationships, then RE F ER EN CES
there may be a potential for development of interventions to alleviate World Medical Association. (2014) “World Medical Association Declaration of
impaired development associated with stunting. The negative Helsinki: Ethical principles for medical research involving human
subjects” Journal of the American College of Dentists 81(3), 14–18.
association between prolonged breastfeeding and development is
Bayley, N. (2006). Bayley scales of infant and toddler development: Bayley‐III.
most likely explained by reverse causality, but this may warrant Harcourt Assessment, Psych. Corporation.
further research. Bernardo, H., & Cesar, V., W. H. Organization (2013). “Long‐term effects of
breastfeeding: A systematic review”.
A U T H O R C O N TR I B U T I O N S Black, M. M., Walker, S. P., Fernald, L. C., Andersen, C. T.,
DiGirolamo, A. M., Lu, C., McCoy, D. C., Fink, G., Shawar, Y. R., &
The authors' contributions were as follows: Henrik Friis, Benedikte
Shiffman, J. (2017). “Early childhood development coming of age:
Grenov, Ezekiel Mupere, Hannah Pesu, Joseph Mbabazi and Rolland Science through the life course”. The Lancet, 389(10064), 77–90.
Mutumba designed the MAGNUS study. Hannah Pesu, Benedikte Briend, A. & Bari, A. (1989). “Breastfeeding improves survival, but not
Grenov, Henrik Friis, Ezekiel Mupere, Joseph Mbabazi and Rolland nutritional status, of 12‐35 months old children in rural Bangladesh”.
European Journal of Clinical Nutrition, 43(9), 603–608.
Mutumba wrote the trial protocol. Henrik Friis and Ezekiel Mupere
Caulfield, L. E., Bentley, M. E., & Ahmed, S. (1996). “Is prolonged breastfeeding
were principal investigators. Joseph Mbabazi, Rolland Mutumba,
associated with malnutrition? Evidence from nineteen demographic and
Hannah Pesu and Benedikte Grenov implemented the study and health surveys”. International Journal of Epidemiology, 25(4), 693–703.
Ezekiel Mupere, Joseph Mbabazi, Hannah Pesu, Rolland Mutumba Choo, Y. Y., Yeleswarapu, S. P., How, C. H., & Agarwal, P. (2019).
and Benedikte Grenov supervised data collection. The statistical “Developmental assessment: practice tips for primary care physi-
cians”. Singapore Medical Journal, 60(2), 57.
analysis plan was developed by Henrik Friis, Christian Ritz and Mette
Cichon, B., Fabiansen, C., Iuel‐Brockdorf, A.‐S., Yaméogo, C. W., Ritz, C.,
F. Olsen. Statistical analysis was performed by Joseph Mbabazi and
Christensen, V. B., Filteau, S., Briend, A., Michaelsen, K. F., & Friis, H.
all coauthors interpreted the results. This manuscript was written by (2018). “Impact of food supplements on hemoglobin, iron status, and
Joseph Mbabazi with input from Mette F. Olsen. All authors critically inflammation in children with moderate acute malnutrition: a 2× 2×
reviewed and approved the final version of the manuscript. 3 factorial randomized trial in Burkina Faso”. The American Journal of
Clinical Nutrition, 107(2), 278–286.
Benedikte Grenov, Joseph Mbabazi, Rolland Mutumba and Henrik
Coates, J., Swindale, A., & Bilinsky, P. (2007). Household food insecurity
Friis accessed and verified the data. Joseph Mbabazi and Mette
access scale (HFIAS) for measurement of household food access:
F. Olsen had final responsibility for the decision to submit the Indicator guide (v.3). (pp. 1–36). Food and Nutrition Technical
manuscript for publication. Assistance Project, Academy for Educational Development.
MBABAZI ET AL. | 11 of 12

Donald, K. A., Wedderburn, C. J., Barnett, W., Nhapi, R. T., Rehman, A. M., “Are early childhood stunting and catch‐up growth associated with
Stadler, J. A., Hoffman, N., Koen, N., Zar, H. J., & Stein, D. J. (2019). school age cognition?—Evidence from an Indian birth cohort”. PLoS
“Risk and protective factors for child development: An observational ONE, 17(3), e0264010.
South African birth cohort”. PLoS Medicine, 16(9), e1002920. Lackey, K. A., Fehrenkamp, B. D., Pace, R. M., Williams, J. E., Meehan, C. L.,
Fink, G., Peet, E., Danaei, G., Andrews, K., McCoy, D. C., Sudfeld, C. R., McGuire, M. A., & McGuire, M. K. (2021). “Breastfeeding beyond
Smith Fawzi, M. C., Ezzati, M., & Fawzi, W. W. (2016). “Schooling and 12 months: is there evidence for health impacts?” Annual Review of
wage income losses due to early‐childhood growth faltering in Nutrition, 41, 283–308.
developing countries: national, regional, and global estimates”. The Leroy, J. L., & Frongillo, E. A. (2019). “Perspective: what does stunting
American Journal of Clinical Nutrition, 104(1), 104–112. really mean? A critical review of the evidence”. Advances in Nutrition,
Ford, N. D. & Stein, A. D. (2016). “Risk factors affecting child cognitive 10(2), 196–204.
development: a summary of nutrition, environment, and Marquis, G. S., Habicht, J.‐P., Lanata, C. F., Black, R. E., & Rasmussen, K. M.
maternal–child interaction indicators for sub‐saharan Africa”. (1997). “Association of breastfeeding and stunting in Peruvian
Journal of Developmental Oorigins of Health and Disease, 7(2), toddlers: an example of reverse causality”. International Journal of
197–217. Epidemiology, 26(2), 349–356.
Gardner, J. M. M., Grantham‐McGregor, S. M., Himes, J., & Chang, S. Mbabazi, J., Pesu, H., Mutumba, R., Filteau, S., Lewis, J. I., Wells, J. C.,
(1999). “Behaviour and development of stunted and nonstunted Olsen, M. F., Briend, A., Michaelsen, K. F., & Mølgaard, C. (2023).
Jamaican children”. The Journal of Child Psychology and Psychiatry and “Effect of milk protein and whey permeate in large quantity lipid‐
Allied Disciplines, 40(5), 819–827. based nutrient supplement on linear growth and body composition
Gladstone, M., Lancaster, G. A., Umar, E., Nyirenda, M., Kayira, E., among stunted children: A randomized 2 × 2 factorial trial in
van den Broek, N. R., & Smyth, R. L. (2010). “The Malawi Uganda”. PLoS Medicine, 20(5), e1004227.
Developmental Assessment Tool (MDAT): the creation, validation, McBride, B. A. & Mills, G. (1993). “A comparison of mother and father
and reliability of a tool to assess child development in rural African involvement with their preschool age children”. Early Childhood
settings”. PLoS Medicine, 7(5), e1000273. Research Quarterly, 8(4), 457–477.
Grummer‐Strawn, L. M. (1993). “Does prolonged breast‐feeding impair Milner, E. M., Kariger, P., Pickering, A. J., Stewart, C. P., Byrd, K., Lin, A.,
child growth? A critical review”. Pediatrics, 91(4), 766–771. Rao, G., Achando, B., Dentz, H. N., & Null, C. (2020). “Association
Hadi, H., Fatimatasari, F., Irwanti, W., Kusuma, C., Alfiana, R. D., between malaria infection and early childhood development
Asshiddiqi, M. I. N., Nugroho, S., Lewis, E. C., & Gittelsohn, J. mediated by anemia in rural Kenya”. International Journal of
(2021). “Exclusive breastfeeding protects young children from Environmental Research and Public Health, 17(3), 902.
stunting in a low‐income population: a study from Eastern MOH. (2009). Policy Guidelines on Infant and Young Child Feeding. January
Indonesia”. Nutrients, 13(12), 4264. 2009. U. Ministry of Health.
Hamadani, J. D., Tofail, F., Hilaly, A., Huda, S. N., Engle, P., & Grantham‐ Mohammed, S., Oakley, L. L., Marston, M., Glynn, J. R., & Calvert,
McGregor, S. M. (2010). “Use of family care indicators and their C. (2022). “The association of breastfeeding with cognitive
relationship with child development in Bangladesh”. Journal of development and educational achievement in sub‐Saharan
Health, Population, and Nutrition, 28(1), 23. Africa: A systematic review”. Journal of Global Health, 12,
Hannah Pesu, R. M., Mbabazi, J., Olsen, M. F., Mølgaard, C., Michaelsen, K. F., 04071.
Ritz, C., Filteau, S., Briend, A., Mupere, E., Friis, H., & Grenov, B. (2021). Mutapi, F., Pfavayi, L., Osakunor, D., Lim, R., Kasambala, M.,
“The role of milk protein and whey permeate in lipid‐based nutrient Mutemeri, A., Rusakaniko, S., Chibanda, D., & Mduluza, T.
supplements on the growth and development of stunted children in (2021). “Assessing early child development and its association
Uganda: A randomized trial protocol (MAGNUS)”. Current Developments with stunting and schistosome infections in rural Zimbabwean
in Nutrition, 81(3), 14–18. children using the griffiths scales of child development”. PLoS
Heesemann, E., Mähler, C., Subramanyam, M. A., & Vollmer, S. (2021). Neglected Tropical Diseases, 15(8), e0009660.
“Pregnancy anaemia, child health and development: A cohort study Mutumba, R., Pesu, H., Mbabazi, J., Greibe, E., Olsen, M. F., Briend, A.,
in rural India”. BMJ Open, 11(11), e046802. Mølgaard, C., Ritz, C., Nabukeera‐Barungi, N., & Mupere, E. (2023).
Hernández‐Luengo, M., Álvarez‐Bueno, C., Martínez‐Hortelano, J. A., “Correlates of iron, cobalamin, folate, and vitamin A status among
Cavero‐Redondo, I., Martínez‐Vizcaíno, V., & Notario‐Pacheco, B. stunted children: A cross‐sectional study in Uganda”. Nutrients,
(2022). “The relationship between breastfeeding and motor devel- 15(15), 3429.
opment in children: a systematic review and meta‐analysis”. Nutrition Nampijja, M., Mutua, A. M., Elliott, A. M., Muriuki, J. M., Abubakar, A.,
Reviews, 80(8), 1821–1835. Webb, E. L., & Atkinson, S. H. (2022). “Low hemoglobin levels are
van den Heuvel, M., Voskuijl, W., Chidzalo, K., Kerac, M., Reijneveld, S. A., associated with reduced psychomotor and language abilities in
Bandsma, R., & Gladstone, M. (2017). “Developmental and beha- young Ugandan children”. Nutrients, 14(7), 1452.
vioural problems in children with severe acute malnutrition in Nyah, M. (2021). “Association between family involvement, play, and literacy
Malawi: A cross–sectional study”. Journal of Global Health, 7(2). skills among preschoolers from low‐income families in Nigeria: the
Hikmahrachim, H. G., & Ronoatmodjo, S. (2020). “Stunting and develop- mediating role of literacy interest”. Turkish Journal of Computer and
mental delays among children aged 6‐59 mo”. International Journal of Mathematics Education (TURCOMAT), 12(10), 4287–4292.
Applied Pharmaceutics, 12(Special Issue 3), 67–71. Olsen, M. F., Iuel‐Brockdorff, A. S., Yaméogo, C. W., Cichon, B.,
Horta, B. L., Bahl, R., Martinés, J. C., & Victora, C. G., W. H. Organization Fabiansen, C., Filteau, S., Phelan, K., Ouédraogo, A., Wells, J. C., &
(2007). “Evidence on the long‐term effects of breastfeeding: systematic Briend, A. (2020). “Early development in children with moderate
review and meta‐analyses”. acute malnutrition: A cross‐sectional study in Burkina Faso”.
Horta, B. L., Loret de Mola, C., & Victora, C. G. (2015). “Breastfeeding and Maternal & child nutrition, 16(2), e12928.
intelligence: a systematic review and meta‐analysis”. Acta Paediatrica, Oumer, A., Fikre, Z., Girum, T., Bedewi, J., Nuriye, K., & Assefa, K. (2022).
104, 14–19. “Stunting and underweight, but not wasting are associated with
ICF, U. B. O. S. U. A. (2017). 2016 Uganda Demographic and Health Survey delay in child development in southwest Ethiopia”. Pediatric Health,
2016 Key Findings. UBOS and ICF. Medicine and Therapeutics, 13, 1.
Koshy, B., Srinivasan, M., Gopalakrishnan, S., Mohan, V. R., Scharf, R., Pereyra‐Elías, R., Quigley, M. A., & Carson, C. (2022). “To what extent
Murray‐Kolb, L., John, S., Beulah, R., Muliyil, J., & Kang, G. (2022). does confounding explain the association between breastfeeding
12 of 12 | MBABAZI ET AL.

duration and cognitive development up to age 14? Findings from the Sirajee, R., Conroy, A. L., Namasopo, S., Opoka, R. O., Lavoie, S., Forgie, S.,
UK Millennium Cohort Study”. PLoS One, 17(5), e0267326. Salami, B. O., & Hawkes, M. T. (2021). “Growth faltering and
Perkins, J. M., Kim, R., Krishna, A., McGovern, M., Aguayo, V. M., & developmental delay in HIV‐exposed uninfected Ugandan infants: a
Subramanian, S. (2017). “Understanding the association between prospective cohort study”. JAIDS Journal of Acquired Immune
stunting and child development in low‐and middle‐income countries: Deficiency Syndromes, 87(1), 730–740.
Next steps for research and intervention”. Social Science & Medicine, Stasinopoulos, M., Rigby, B., Voudouris, V., Akantziliotou, C., Enea, M., &
193, 101–109. Kiose, D. (2022). Package ‘gamlss’.
Poh, B. K., Rojroonwasinkul, N., Le Nyugen, B. K., Budiman, B., Ng, L. O., Syeda, B., Agho, K., Wilson, L., Maheshwari, G. K., & Raza, M. Q. (2021).
Soonthorndhada, K., Xuyen, H. T., Deurenberg, P., & Parikh, P. “Relationship between breastfeeding duration and undernutrition
(2013). “Relationship between anthropometric indicators and cogni- conditions among children aged 0–3 years in Pakistan”. International
tive performance in southeast asian school‐aged children”. British Journal of Pediatrics and Adolescent Medicine, 8(1), 10–17.
Journal of Nutrition, 110(S3), S57–S64. UNICEF, World Bank. (2023). Joint Child Malnutrition Estimates: Key
Prado, E. L., Abbeddou, S., Adu‐Afarwuah, S., Arimond, M., Ashorn, P., findings of the 2023 edition, Levels and trends in child mal-
Ashorn, U., Bendabenda, J., Brown, K. H., Hess, S. Y., & nutrition. WHO.
Kortekangas, E. (2017). “Predictors and pathways of language and WHO. (2017a). Global Nutrition Monitoring Framework: Operational
motor development in four prospective cohorts of young children in guidance for tracking progress in meeting targets for 2025. W. H.
Ghana, Malawi, and Burkina Faso”. Journal of Child Psychology and Organization.
Psychiatry, 58(11), 1264–1275. WHO. (2017b). “Stunted growth and development: context, causes and
Quinn, P. J., O'Callaghan, M., Williams, G. M., Najman, J. M., Andersen, M., consequences”, World Health Organization.
& Bor, W. (2001). The effect of breastfeeding on child development
at 5 years: a cohort study. Journal of Paediatrics and Child Health,
37(5), 465–469. SUPP ORTING INFO RM ATION
Rosyidah, M., Dewi, Y. L. R., & Qadrijati, I. (2021). “Effects of stunting on Additional supporting information can be found online in the
child development: A meta‐analysis”. Journal of Maternal and Child
Supporting Information section at the end of this article.
Health, 6(1), 25–34.
Sacker, A., Quigley, M. A., & Kelly, Y. J. (2006). “Breastfeeding and
developmental delay: findings from the millennium cohort study”.
Pediatrics, 118(3), e682–e689.
Scarpa, G., Berrang‐Ford, L., Galazoula, M., Kakwangire, P., Namanya, How to cite this article: Mbabazi, J., Pesu, H., Mutumba, R.,
D. B., Tushemerirwe, F., Ahumuza, L., & Cade, J. E. (2022). “Identifying Bromley, K., Ritz, C., Filteau, S., Briend, A., Mupere, E.,
predictors for minimum dietary diversity and minimum meal frequency in Grenov, B., Friis, H., & Olsen, M. F. (2024). Correlates of early
children aged 6–23 months in Uganda”. Nutrients, 14(24), 5208.
child development among children with stunting:
Scharf, R. J., Rogawski, E. T., Murray‐Kolb, L. E., Maphula, A., Svensen, E.,
Tofail, F., Rasheed, M., Abreu, C., Vasquez, A. O., & Shrestha, R. A cross‐sectional study in Uganda. Maternal & Child
(2018). “Early childhood growth and cognitive outcomes: findings Nutrition, 20, e13619. https://doi.org/10.1111/mcn.13619
from the MAL‐ED study”. Maternal & Child Nutrition, 14(3), e12584.