Detection of fetal malnutrition at birth using clinical assessment of nutritional status (CAN) score and anthropometric indices and its associated risk factors among term newborns in southern Ethiopia: facility-based cross-sectional study

Abstract

Objectives

This study aimed to determine the prevalence of fetal malnutrition (FM) among term singleton newborns at birth using clinical assessment of nutritional status (CAN) score. The study also compared the CAN score with selected anthropometric indices to describe their ability to identify FM. Furthermore, it sought to identify maternal and neonatal factors linked with FM among term infants delivered in four public hospitals in South Ethiopia.

Design

Cross-sectional study.

Setting

The study was conducted at four public hospitals in South Ethiopia Regional State.

Participants

480 live-born full-term babies and their mothers delivered at four public hospitals between 6 June 2021 and 19 April 2022.

Outcome measures

FM status at birth was the outcome variable. FM was assessed using both the CAN score and selected anthropometric indices. At birth, measurements of height, weight, head circumference (HC) and mid-arm circumference (MAC) were taken. The MAC/HC ratio, body mass index (BMI) and ponderal index (PI) were computed and contrasted with reference curves. By examining and calculating subcutaneous tissue and muscle loss, the CAN score was used to determine the newborn’s CAN status. A CAN score of <25 was considered FM.

Results

This study of 480 newborns found varying rates of FM depending on the diagnostic criteria used. The prevalence of FM was highest when assessed by CAN score (26.04%), followed by PI (17.29%), BMI (16.25%), MAC/HC ratio (13.54%) and birth size-for-gestational-age (11.25%). Not attending formal education (AOR 3.31 95% CI 1.14 to 9.64), age between 25–29 years (AOR 2.66 95% CI 1.19 to 5.98), primi-parity (AOR 9.63 95% CI 1.42 to 65.46), having <= two antenatal clinic (ANC) visits (AOR 2.34 95% CI 1.05 to 5.24) and pregnancy-induced hypertension (PIH) (AOR 16.99 95% CI 4.20 to 68.81) were maternal factors significantly associated with increased risk of FM.

Conclusions

FM was prevalent among term neonates, with the CAN score showing higher prevalence than anthropometric indices. Education level of the mother, primi-parity, fewer ANC visits, maternal age between 25–29 years and PIH were significant risk factors. FM might be most effectively recognised using the CAN score, which can be done without advanced tools.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Comprehensive assessment of newborn growth and nutritional status using both clinical evaluation (clinical assessment of nutritional status score) and standardised anthropometric indices.

• Use of sex-specific Intergrowth-21st standards to classify birth weight-for-gestational age, ensuring appropriate growth status evaluation.

• Potential bias in gestational age estimation due to reliance on maternal recall of last menstrual period and limitations in ultrasound availability and quick assessment techniques.

• Subjectivity in neonatal evaluation using scoring systems, which may introduce variability in assessment outcomes.

• The inability to assess maternal prepregnancy body mass index, as only early postpartum weight was available, potentially affecting the accuracy of maternal nutritional status evaluation.

Introduction

Fetal malnutrition (FM) is a major public health issue worldwide. It has been estimated that between 2% and 10% of the total number of births worldwide experience FM, and the prevalence of FM is higher in developing countries.¹ In Africa, the prevalence of FM in newborns, based on the clinical assessment of nutritional status (CAN) score, varies from 13.5% in Nigeria² to 30% in Egypt,³ while in Ethiopia the prevalence of FM varies between 12.3% and 21.7%.^{4 5} FM adversely affects neonatal growth and development, increases vulnerability to infections and is linked with short- and long-term complications, such as neurodevelopmental sequelae and the risk of chronic diseases in the future.⁶

The term FM, which was created by Scott and Usher,⁷ refers to a state of undernutrition in utero resulting from inadequate supply and/or utilisation of nutrients.⁸ FM is defined as a clinical state of a baby of any birth weight and gestational age characterised by obvious intrauterine loss of or failure to acquire normal amounts of subcutaneous fat and muscle mass as it experienced deprivation of adequate calories, proteins and other nutrients necessary for proper growth and development.⁸ Infants with FM may be categorised as little, low birth weight (LBW) or may be appropriate-for-gestational-age (AGA).⁹ Consequently, all newborns should be assessed for FM, regardless of birth weight or gestational age.¹⁰

Failure to properly assess neonatal nutritional status at birth can result in nutritional disorders that negatively impact infant development and health.11,13 Addressing FM is essential due to its potential to harm multiple organ systems. Studies indicate that 39% of severely malnourished infants develop intellectual and neurological impairments.^{14 15} Children with FM are more likely to have lower IQ scores, require special education or experience neurological disabilities, learning disorders or seizures later in life. FM increases the risk of high-risk deliveries and perinatal complications.1016,18

In the clinical and research arena, various anthropometric parameters have been used to assess and diagnose FM. These parameters include ponderal index (PI), body mass index (BMI) and the ratio of mid-upper arm circumference to head circumference (MAC/HC). Although these parameters are objective and easy to measure, their sensitivity and specificity are limited in comparison with the clinical assessment.¹⁹ The CAN score assesses physical signs of tissue loss and is an objective measure of nutritional status at birth.¹⁹ The CAN score is a simple and inexpensive clinical index for identifying FM among term newborns and has been shown to be a more reliable indicator than most anthropometric approaches.^{20 21} In Ethiopia, as well as other sub-Saharan African countries, there is little evidence on the prevalence of FM and its associated risk factors, despite there being a high population projection and the risk of malnutrition. It is also crucial to attain the Sustainable Development Goal of newborn health, which targets to reduce neonatal mortality to at least down to 12 deaths per 1000 live births.²² Therefore, the main objective of this study was to determine the prevalence of FM among term singleton newborns using CAN score and selected anthropometric indices and identify risk factors associated with FM in four selected public hospitals in South Ethiopia Regional State.

Methods and materials

Study design and setting

A hospital-based cross-sectional study was conducted in South Ethiopia Regional State. In the Region, four Zones were considered. This study was based at four Public Hospitals (one in each Zone): Arba Minch General Hospital in Gamo Zone, Sawula General Hospital in Gofa Zone, Karat General Hospital in Konso Zone and Jinka General Hospital in Ari Zone. General hospitals provide inpatient and outpatient services, including maternal and child care. They serve as a referral centre for primary hospitals. Hence, the setups of general hospitals are similar in kind of services they deliver, health information recording system and in number and qualification of health professionals required.

Participants and inclusion criteria

All live full-term newborns delivered in the study hospitals between 6 June 2021 and 19 April 2022, were considered eligible for the study. The inclusion criteria comprised inborn, clinically stable and singleton newborns delivered consecutively in the study hospitals. Newborns were excluded from the study if they had congenital anomalies, required neonatal intensive care unit (NICU) care, or were born to mothers with an unreliable estimation of gestational age.¹⁷ Congenital anomalies were specifically excluded because they can change the body proportions and physical features of the baby, thereby affecting the assessment of nutritional status of the fetus.

Sample size determination and sampling procedures

Using EPI-INFO V.7 stat calc programme, sample size was determined using single population proportion formula by considering the following assumptions: 95% confidence level and 3.5% margin of error. Taking 16.7% prevalence of FM at birth based on previous study in Nigeria,²³ the minimum required sample size for the study was 436. Considering a 10% non-response rate, the final sample size becomes 480 term newborns.

The number of deliveries conducted at each hospital over the past three consecutive months was used as a reference. The sample size for each hospital was then allocated proportionally based on the number of skilled deliveries attended during the last 3 months of 2020. Mother-newborn pairs residing in the locality were consecutively recruited from the delivery and labour wards immediately after delivery throughout the study period.

Study variables and definitions

Outcome variable

The primary outcome was FM status, assessed using CAN score in addition to four other anthropometric indices. The CAN score was determined within 48 hours of birth. It evaluates nine readily observable signs of newborn malnutrition: loss of subcutaneous tissue and muscle mass in specific areas (hair, cheeks, neck and chin, arms, back, buttocks, legs, chest and abdomen). Each area was visually inspected and scored on a scale of 1 to 4; 1: maximum evidence of malnutrition. 4: evidence of good nutrition. The total CAN score ranges from 9 (lowest) to 36 (highest). A score below 25 is indicative of FM.^{23 24} Small-for-gestational age (SGA) is defined as birth weight below the 10th percentile for gestational age, adjusted for sex, compared with a standardised reference population.²⁵ PI was calculated as=Weight (g)×100/Length (cm)³. A PI value below 2.2 g/cm³ was considered indicative of malnutrition.²⁶ MAC/HC ratio: a cut-off value of 0.27 was used to define malnutrition.²⁷ BMI calculated as=Weight (kg)/Length (m)² at birth and plotted on the Brock et al newborn BMI curve.²⁸ A cut-off value of 11.20 kg/m² was considered indicative of malnutrition.²⁸

Explanatory variables

Categorised into three groups: sociodemographic, maternal and obstetric-related and neonatal-related factors.

Sociodemographic characteristics

Marital status of mother, occupation status of the mother, level of education of the mother, residence, religion, age at first marriage and birth, monthly reported income, current age of mother, family size.

Maternal and obstetric-related characteristics

Weight (kg), height (m), maternal mid-upper arm circumference (MUAC), history of abortion, history of stillbirth, dates of antenatal clinic (ANC) booking, number of ANC visits, gravidity (no. of pregnancy), parity (number of birth), number of living children, date of last menstrual period (LMP), fetal presentation, mode of delivery, diagnosed medical conditions (like pregnancy-induced hypertension (PIH), chronic hypertension, urinary tract infection (UTI), diabetes mellitus (DM), HIV/AIDS, anaemia), types of obstetric (pregnancy/labour) complications (antepartum haemorrhage (APH), breech position, multiple pregnancy, pre-eclampsia/eclampsia, premature rupture of membranes (PROM), prolonged/obstructed labour, fetal distress, post partum haemorrhage (PPH), past medical, obstetric and surgical histories; presence or otherwise of proteinuria, use of long-lasting insecticide-treated bed nets (ITNs), history of adverse birth outcome.

Newborn-related characteristics

Birth weight (g), weight (g), HC, MUAC and supine length, birth interval, birth order, other biodata such as sex of the newborn, 1- and 5 min Apgar scores, gestational age at birth, date of birth, age on admission, date of death if the newborn has died, intrauterine growth status at birth. Birthweight-for-gestational-age centiles were calculated for each live-born singleton newborn using sex-specific Intergrowth-21st newborn size standards.²⁹ Based on these standards, newborns weighing below the 10th percentile were classified as SGA, those between the 10th and 90th percentiles as AGA, and those above the 90th percentile as large for gestational age (LGA).²⁹

Data collection instruments and process

Data were collected using data extraction checklist, interviewer-administered structured questionnaire and direct physical examination. Maternal and neonatal clinical information was extracted from hospital medical records using a data extraction checklist. Maternal variables included weight (kg), height (m), maternal MUAC, date of LMP, fetal presentation, mode of delivery and diagnosed medical conditions such as PIH, chronic hypertension, UTI, DM, HIV/AIDS and anaemia. Information on obstetric or pregnancy-related complications, including APH, breech position, multiple pregnancy, pre-eclampsia/eclampsia, PROM, prolonged/obstructed labour, fetal distress and PPH, was also extracted. Neonatal information obtained from medical records included birth weight (g), weight (g), HC, MUAC, supine length, sex of the newborn, first- and fifth-minute Apgar scores, date of birth, age on admission and date of death if died. Additional maternal and reproductive history variables such as history of abortion, history of stillbirth, dates of ANC booking, number of ANC visits, gravidity, parity, number of living children, past medical, obstetric and surgical histories; presence of proteinuria, use of ITNs, history of adverse birth outcome were collected through interviews with mothers or caregivers using structured questionnaire. Sociodemographic characteristics, such as maternal age, educational status, occupation, residence and marital status, were also obtained by interviewing mothers or caregivers.

At each site, three trained health professionals were responsible for collecting the required data. Interviews with mothers/caregivers were conducted after birth but before discharge from health facilities. Gestational age was estimated using LMP or, when available, early ultrasound readings. Birth weight was measured to the nearest 10 g on an electronic scale while the infant was unclothed. Length was measured to the nearest 0.1 cm using an infant measuring board. HC was measured at the maximum occipital-frontal circumference to the nearest 0.1 cm using a non-stretchable measuring tape. MAC was measured on the left arm, midway between the acromion and the olecranon process, to the nearest 0.1 cm using a non-stretchable measuring tape. Furthermore, each newborn was clinically evaluated through direct physical examination within 2 days of birth for signs of malnutrition using the CAN score, as adapted by Metcoff.²⁴

Data management and statistical analysis

Data were collected using KoBo Toolbox, V.2.015 data management system. After screening (through frequent check-up and cleaning through running simple frequencies and cross-tabulations) and correction of data entry errors, data were exported to STATA V.14 statistical software for analysis. Data were described using standard statistical methods. Categorical variables were described using frequency and percentage, while continuous variables were described using mean and SD. Sex-specific birthweight-for-gestational-age percentiles and length-for-gestational-age percentiles were produced for each live-born singleton newborn, using sex-specific Intergrowth-21st newborn size standards.²⁹ Both bivariate and multivariable logistic regression analyses were performed to identify factors associated with FM defined by the CAN score. The CAN score was used as the primary outcome measure because it is a validated clinical tool that detects FM based on subcutaneous fat and muscle wasting, independent of birth weight and gestational age.¹⁷ Those independent variables that were significant in bivariate logistic regression at 25% were eligible for multivariable logistic regression. Maternal educational status, age of mother, gravidity, parity, number of ANC visits, PIH, sex of the newborn and newborn birth size-for-gestational age were the variables eligible for multivariable logistic regression. Backward stepwise regression was used to select variables. In the final model, there were eight variables included as adjusted factors. The Hosmer-Lemeshow test was used to test the goodness of fit for the model. Multicollinearity was checked using variance inflation factors (>10 taken as the existence of collinearity). The variables selected for the model did not show a multicollinearity problem. Extent of strength was presented using ORs and 95% CIs. P value <0.05 was used as a cut point to determine significance in multiple logistic regression. The results were presented using texts, tables, charts and graphs.

Patient and public involvement

Parents or the public were not directly involved in the design, conduct or plans for the dissemination of our research.

Results

Sociodemographic characteristics of study participants

A total of 480 full-term, singleton, live-born newborns with a mean gestational age of 39.49±1.45 weeks were included in the study. Out of these, 230 (47.92%) were boys and 250 (52.08%) were girls. Table 1 depicts the overall summary of anthropometric parameters and comparison of anthropometric and CAN score data between neonates with and without FM. The overall mean anthropometric values were as follows: birth weight 3360.35±506.87 g, length 49.83±2.07 cm, HC 33.24±2.29 cm, MUAC 10.37±1.41 cm, PI 2.72±0.39, MAC/HC ratio 0.31±0.04, BMI 13.53±1.79 and CAN score 26.34±3.61. Neonates with FM had statistically significant lower average values of birth weight, HC, MUAC, PI, MAC/HC ratio, BMI and CAN scores compared with those without FM (p<0.04). Their birth length was also a bit lower, but the difference wasn’t statistically significant.

Table 1. The overall summary of anthropometric parameters and comparison of anthropometric data of term newborns with and without fetal malnutrition (mean±SD).

Anthropometric parameters	Nutritional status (mean±SD)			Statistical values
	Babies with FM (n=125)	Babies without FM (n=355)	Over all total mean±SD	t-test	P value
Birth weight (g)	3066.16±490.01	3463.94±471.38	3360.35±506.87	−8.03	0.001
Birth length (cm)	49.78±1.94	49.84±2.11	49.83±2.07	−0.27	0.39
Head circumference (cm)	32.84±2.55	33.381±2.18	33.24±2.29	−2.307	0.012
MAC (cm)	9.95±1.33	10.51±1.40	10.37±1.41	−3.916	0.001
Ponderal index	2.49±0.45	2.799±0.33	2.72±0.39	−8.00	0.001
MAC/HC ratio	0.30±0.04	0.32±0.04	0.31±0.04	−2.67	0.04
Body mass index	12.39±2.02	13.92±1.51	13.53±1.79	−8.86	0.001
CAN score	21.82±2.52	27.93±2.09	26.34±3.61	−24.26	0.001

CAN, clinical assessment of nutritional status; FM, fetal malnutrition; MAC/HC, mid-arm/head circumference.

Selected maternal characteristics of the study population

Of the 480 pairs included, 108 (22.50%) were recruited from Arba Minch general hospital, 145 (30.21%) from Jinka hospital, 71 (14.79%) from Karat hospital and 156 (32.50%) from Sawla general hospital. Of the mother, 45.63% were rural residents while the remaining 261 (54.37%) were urban residents. Almost all 442 (92.08%) of the mothers were married, while only 31 (6.46%) were single and 7 (1.46%) were divorced/widowed (table 2). Majority of the mothers, 246 (51.25%) were housewives by occupation, followed by government workers, 100 (20.83%). About 119 (24.79%) of the mothers had no education, and 134 (27.92%) of the mothers had completed primary school (grades 1 to 8). Closely, 94 (19.58%) of the mothers had completed secondary school (grade 9 to 12) while 133 (27.71%) were above secondary.

Table 2. Sociodemographic characteristics of neonates of mothers included in the study of clinical assessment of nutritional status of term newborns in southern Ethiopia.

Characteristics	Malnourished, n (%)	Well-nourished, n (%)	Total, n (%)
Residence
Rural	59 (12.29)	160 (33.33)	219 (45.63)
Urban	66 (13.75)	195 (40.63)	261 (54.37)
Marital status
Married	116 (24.17)	326 (67.92)	442 (92.08)
Single	8 (1.67)	23 (4.79)	31 (6.46)
Widowed/divorced	1 (0.21)	6 (1.25)	7 (1.46)
Occupational status of mother
Farmer	19 (3.96)	48 (10.00)	67 (13.96)
Government worker	23 (4.79)	77 (16.04)	100 (20.83)
Housewife	65 (13.54)	181 (37.71)	246 (51.25)
Merchant	8 (1.67)	14 (2.92)	22 (4.58)
Private worker	10 (2.08)	35 (7.29)	45 (9.38)
Level of education
No formal education	44 (9.17)	75 (15.63)	119 (24.79)
Primary	35 (7.29)	99 (20.63)	134 (27.92)
Secondary	16 (3.33)	78 (16.25)	94 (19.58)
College	30 (6.25)	103 (21.46)	133 (27.71)

Prevalence of fetal malnutrition

The distribution of the study participants into well-nourished and malnourished groups, based on different methods, is shown in table 3. The prevalence of FM as detected by the CAN score was 125 (26.04%) (95% CI 22.17% to 30.21%), while the PI identified 83 (17.29%) (95% CI 14.01% to 20.98%) of participants as malnourished. BMI detected 78 (16.25%) of the study participants as malnourished. Among the 83 (17.29%) newborns classified as malnourished by the PI, the CAN score identified 17 (20.48%) as clinically well-nourished and 59 (14.86%) as malnourished. When classified by BMI, 78 (16.25%) (95% CI 14.01% to 20.98%) of the newborns were identified as malnourished. However, using the CAN score, 17 (21.79%) of those classified as malnourished by BMI were found to be clinically well-nourished. In contrast, among the 402 (83.75%) newborns with normal BMI, 64 (15.92%) were identified as malnourished based on the CAN score. These findings were statistically significant (table 3).

Table 3. Comparison of fetal malnutrition detection by anthropometric indices (PI, BMI and MAC/HC) and birth size-for-gestational age with the CAN score among term newborns.

Anthropometric variables	CAN score (%)			χ2 test	P value
	Malnourished (n=125)	Well-nourished (n=355)	Total (n=480)
PI (ponderal index)
Malnourished	66 (52.80)	17 (4.79)	83 (17.29)	149.01	0.001
Well-nourished	59 (47.20)	338 (95.21)	397 (82.71)
BMI (body mass index)
Malnourished	61 (48.80)	17 (4.79)	78 (16.25)	131.57	0.001
Well-nourished	64 (51.20)	338 (95.21)	402 (83.75)
MAC/HC ratio
Malnourished	26 (20.80)	39 (10.99)	65 (13.54)	7.605	0.006
Well-nourished	99 (79.20)	316 (89.01)	415 (86.46)
Birthweight-for-gestational age
SGA	46 (36.80)	8 (2.25)	54 (11.25)	110.84	0.001
AGA+LGA	79 (63.20)	347 (97.75)	426 (88.75)

AGA, appropriate-for-gestational-age; CAN, clinical assessment of nutritional status; LGA, large for gestational age; MAC/HC, mid-arm/head circumference; SGA, small for gestational age.

Regarding the MAC/HC ratio, 65 (13.54%) newborns (95% CI 10.61% to 16.93%) were identified as malnourished. Among these 65 newborns, 39 (60.00% of the newborns) were classified as well-nourished by CAN score, while 99 (23.86%) of them were clinically malnourished (table 3). According to the Intergrowth-21st newborn growth standard, 315 newborns (65.63%, 95% CI 61.19% to 69.87%) were classified as AGA, 111 (23.13%, 95% CI 19.42% to 27.16%) were LGA and 54 (11.25%, 95% CI 8.57% to 14.42%) were classified as SGA. The CAN score identified 46 (85.19%) of SGA babies, 63 (20%) of AGA babies and 16 (14.41%) of LGA babies as malnourished. Among SGA newborns, 8 (14.81%) were identified as well-nourished according to the CAN score, whereas 347 (97.75%) of AGA and LGA newborns were classified as well-nourished. Thus, 8 (14.81%) of SGA babies were small but not malnourished, and 79 (18.55%) of AGA and LGA babies were fully grown but malnourished. The association between birth size for gestational age and FM was statistically significant (table 3).

Risk factors associated with fetal malnutrition

In the multivariable logistic regression, mothers who have no formal education, age of mother in the 25–29 years, primiparous mothers, having <= two ANC visits, PIH, being female and being SGA babies were significantly associated with increased risk of FM (table 4). The model demonstrated a good fit according to Hosmer-Lemeshow goodness-of-fit test (χ²=13.27, p=0.10) and explained approximately R²=0.79 of the variability in FM. Accordingly, the odds of having FM were three times higher among mothers who had not attended formal education compared with mothers who attended college education (AOR 3.31 95% CI 1.14 to 9.64). Mothers of newborns whose age was 25 to 29 years had an increased risk of FM compared with mothers aged 30 to 34 years (AOR 2.66 95% CI 1.19 to 5.98). Primiparous mothers, mothers who attended ≤ two ANC visits and mothers who had PIH were nine times (AOR 9.63 95% CI 1.42 to 65.46), two times (AOR 2.34 95% CI 1.05 to 5.24) and 17 times (AOR 16.99 95% CI 4.20 to 68.81) more likely to have FM compared with multiparous (III) mothers, mothers who attended ≥ four ANC visits and who did not have PIH, respectively. Female fetuses were nearly two times more likely to be exposed to FM than male counterparts (AOR 1.93 95% CI 1.12 to 3.30). Newborns delivered at less than the 10th percentile (SGA) were 30 times more likely to be exposed to FM than those born appropriate birth size for gestational age (AOR 30.68 95% CI 12.42 to 75.77) (table 4).

Table 4. Binary and multivariable logistic regression showing factors associated with fetal malnutrition at birth in four selected public hospitals, southern Ethiopia, 2022.

Characteristics	COR (95% CI for COR)	P value	AOR (95% CI for AOR)	P value
Maternal educational level
No formal education	2.01 (1.16 to 3.49)	0.013	3.31 (1.14 to 9.64)	0.028
Primary	1.21 (0.69 to 2.13)	0.49	1.44 (0.54 to 3.82)	0.463
Secondary	0.70 (0.36 to 1.38)	0.31	0.81 (0.30 to 2.18)	0.671
College (ref.)	1		1
Mother occupation
Farmer	1.33 (0.65 to 2.69)	0.435	0.65 (0.19 to 2.25)	0.499
Housewife	1.20 (0.70 to 2.07)	0.508	0.65 (0.23 to 1.82)	0.411
Merchant	1.91 (0.71 to 5.13)	0.197	2.38 (0.63 to 8.91)	0.199
Private worker	0.96 (0.41 to 2.22)	0.918	1.03 (0.35 to 3.08)	0.954
Government worker (ref.)	1	1	1
Age of mother (years)
15–24	1.18 (0.66 to 2.11)	0.582	1.84 (0.70 to 4.82)	0.216
25–29	0.94 (0.53 to 1.66)	0.25	2.66 (1.19 to 5.98)	0.018
≥35	0.65 (0.22 to 1.92)	0.438	0.54 (0.12 to 2.44)	0.422
30–34 (ref.)	1		1
Gravidity
Primigravida	1.45 (0.85 to 2.47)	0.171	0.21 (0.03 to 1.40)	0.107
Gravida II	0.55 (0.29 to 1.05)	0.07	0.46 (0.14 to 1.58)	0.217
Gravida IV and above	0.89 (0.49 to 1.62)	0.71	0.90 (0.22 to 3.73)	0.882
Gravida III (ref.)	1		1
Parity
Primiparous	1.73 (1.01 to 2.96)	0.047	9.63 (1.42 to 65.46)	0.021
Multiparous (II)	0.79 (0.42 to 1.46)	0.451	2.32 (0.71 to 7.57)	0.165
Multiparous (4+)	1.12 (0.59 to 2.14)	0.721	2.10 (0.48 to 9.27)	0.326
Multiparous (III) (ref.)	1		1
History of stillbirth
Yes	1.25 (0.71 to 2.20)	0.444	1.42 (0.61 to 3.26)	0.414
No (ref.)	1		1
Height of mother
<150 cm	1.24 (0.61 to 2.52)	0.552	0.82 (0.31 to 2.15)	0.69
≥150 cm (ref.)	1		1
Number of ANC visits
≤two visits	3.94 (2.14 to 7.24)	0.001	2.34 (1.05 to 5.24)	0.038
Three visits	1.29 (0.80 to 2.10)	0.296	1.28 (0.71 to 2.30)	0.412
≥four visits (ref.)	1		1
Pregnancy-induced hypertension
Yes	4.32 (1.61 to 11.62)	0.004	16.99 (4.20 to 68.81)	0.001
No (ref.)	1		1
Maternal chronic hypertension
Yes	0.95 (0.19 to 4.75)	0.946	2.94 (0.42 to 20.68)	0.279
No (ref.)	1		1
Maternal urinary tract infection (UTI)
Yes	0.81 (0.26 to 2.49)	0.707	0.61 (0.13 to 2.93)	0.536
No (ref.)	1		1
Maternal DM
Yes	0.46 (0.10 to 2.11)	0.32	0.32 (0.04 to 2.48)	0.278
No (ref.)	1		1
Maternal anaemia
Yes	1.14 (0.43 to 3.01)	0.787	1.00 (0.25 to 4.05)	0.997
No (ref.)	1		1
Sex of the newborn
Female	1.09 (0.72 to 1.63)	0.25	1.93 (1.12 to 3.30)	0.017
Male (ref.)	1		1
Newborn birth size for gestational age
SGA	23.00 (10.34 to 51.18)	0.001	30.68 (12.42 to 75.77)	0.001
LGA	0.67 (0.37 to 1.22)	0.195	0.76 (0.39 to 1.48)	0.416
AGA (ref.)	1		1

AGA, appropriate-for-gestational-age; ANC, antenatal clinic; DM, diabetes mellitus; LGA, large for gestational age; SGA, small for gestational age.

Discussion

This facility-based cross-sectional study included 480 singleton term newborns. The CAN score, a method for detecting FM based on physical characteristics, was used. A prevalence of 26.04% (85.19% of SGA, 20% of AGA) of FM was observed among term newborns. These findings are consistent with previous studies conducted in Ethiopia (21.7%,⁵ Nigeria (18.8%)³⁰ and in India (17.5%,¹¹ 21.2%²⁰ and 27.97%,¹⁰ as well as Iran (24%).¹⁷ This consistency may be attributed to the shared characteristics of these studies: a focus on term newborns, the use of CAN score as the screening tool, and uniform criteria for defining FM.³¹ Nevertheless, the prevalence observed in this study was higher than that reported in Spain (7.6%).²¹ The variation of results among various countries may be due to lack of adequate maternal healthcare and nutrition in resource-limiting countries as compared with resource-rich countries.

In Ethiopia, there is limited literature on prevalence of FM. Traditionally, poor fetal nutritional status has been inferred using measures such as LBW and SGA; however, these indicators are insufficient for accurately assessing FM. Previous studies have examined FM prevalence across different gestational age categories using both the CAN score and anthropometric measures, emphasising that birth weight alone does not reliably reflect fetal nutritional status.^{2 32} It is well-established that birth weight alone may not accurately reflect fetal nutritional status. Earlier studies^{21 24} have demonstrated that not all SGA babies exhibit FM, and conversely, some AGA babies may display features of FM. Metcoff²⁴ reported FM prevalence of 5.5% in AGA newborns and 54.0% in SGA births. Similarly, Sankhyan et al¹⁰ identified FM in 57.1% of SGA and 3.8% of AGA newborns. Adebami et al¹⁵ found FM in 10.6% of term AGA babies using CAN score. In this study, 20% of term AGA newborns exhibited FM based on CAN score, highlighting the presence of malnutrition despite normal birth weight. Among SGA newborns, 85.19% were malnourished, underscoring the distinction between SGA and FM.

While SGA increases the likelihood of FM,^{15 24} it is crucial to recognise that these conditions are not synonymous. The CAN score proved effective in precisely detecting malnourished term newborns, aligning with Metcoff’s observation²⁴ that not all SGA babies exhibit FM. Therefore, comprehensive clinical assessments that include evaluations of hair texture, subcutaneous fat, muscle mass and other physical characteristics are essential to accurately determine FM status.

Anthropometric indices are essential for assessing intrauterine growth and establishing a baseline for tracking the nutritional status of newborns. For determining the nutritional condition of newborns, researchers have proposed that analysing body proportions offers more insightful information than depending only on age-based metrics.^{10 11 30 33}

According to PI, the prevalence of FM in this study was 17.29%. This finding is higher than those found in Nigerian studies (8.1%¹⁵ and 10.3%²) but lower than the Kashyap and Dwivedi³³ study in India (27.8%). The inclusion of preterm infants (≥35 weeks) in Kashyap and Dwivedi’s analysis may be the reason for the higher prevalence they reported.³³

In this study, the mean PI of babies with FM was relatively low (2.49±0.45) compared with babies without FM (2.799±0.33). Low PI means low weight-for-length, also called “thinness”, a term that does not essentially indicate the pathological process called wasting. A baby may be thin from the effects of chronic dietary deficit. Though around 2%–3% of normal newborns might show thinness with low PI without inevitably being wasted.³⁴ Interestingly, some newborns with normal PI were identified as malnourished using the CAN score. This finding underscores the limitations of relying solely on PI for assessing FM. One possible explanation is that FM can also affect the infant’s length, thereby influencing the PI measurement and reducing its accuracy as a standalone diagnostic tool.

The prevalence of FM using MAC/HC ratio was lower, at 13.54%, compared with 49.76% reported in an Indian study.³⁵ This discrepancy may be due to differences in cut-off values used for detecting FM. Curiously, not all newborns identified as malnourished by anthropometric indices were classified as such by the CAN score, and vice versa. Consistent with previous studies,^{17 36} the CAN score identified more cases of FM than anthropometric indices, reinforcing its reliability as a diagnostic tool for detecting FM.

In this study, the following maternal risk factors were significantly linked to FM: mothers who have not attended formal education, ages between 25–29 years, primiparous mothers, having <= two ANC visits, PIH. Neonatal factors linked with a higher risk of FM were female sex and being born SGA. Adebami et al³⁷ similarly reported that lack of antenatal care, young maternal age (<18 years), primiparous women and inadequate prenatal care were all substantially linked to higher risk of FM. Notably, PIH emerged as a significant pregnancy-related factor in this study (AOR 16.99 95% CI 4.20 to 68.81), aligning with the findings of Adebami et al³⁷ who also highlighted its adverse impact on fetal nutrition. Hypertension and pre-eclampsia bring fibrinoid disintegration of arteriolar media with ensuing intimal stiffening and decreased luminal diameter.³⁸ It causes reduced permeability of essential nutrient across the placental barrier to the fetus and has, therefore, been implicated as a causative factor of FM.³⁹ These results add to the evidence implicating PIH as a threat to fetal well-being. However, other common infections during pregnancy like UTI, malaria, anaemia and maternal DM were not linked to FM in this study.

There are multiple limitations to this study. Due to the dependence on maternal memory of LMP, as well as limitations in ultrasound estimation and quick gestational age assessment techniques, there may be bias in the estimation of gestational age. The assessment may also be interpreted subjectively, as is the case with various neonatal evaluation score systems. As reported by Adebami et al¹⁸ and Soundarya et al,¹⁷ it was anticipated that maternal BMI, a crucial measure of maternal nutritional status, would affect newborn nutritional status. However, weight increase and fluid retention during pregnancy affect BMI, which was determined in this study using the mother’s early postpartum weight. However, there was no prepregnancy weight or BMI available, which would have yielded more precise measurements. Though anthropometric measures like PI, BMI, MAC/HC ratio and birth size for gestational age are widely used for diagnosing FM, each has its own limitations. All these anthropometric measures are based on measurements obtained from body and may not account for soft tissue depletion or loss of subcutaneous fat tissue. In this regard, the CAN score can be used for diagnosing FM in normal birth weight or normal-sized for gestational age babies. However, this test is subjective in nature and depends on the examiner’s expertise, which can lead to inter-observer variability. The study uses a cross-sectional design, which might limit the generalisability of the findings. Moreover, the results show association and not causation.

Conclusions

In conclusion, FM was common among term neonates, with the CAN score identifying a higher prevalence than anthropometric indices. Newborns with FM are high-risk and require close monitoring and proactive management to improve postnatal outcomes. Maternal factors such as inadequate antenatal care, maternal age of 25–29 years, primiparity, pregnancy-induced PIH and infections like UTIs were significantly associated with FM. FM might be most effectively recognised using the CAN score, which can be done without advanced tools. Strengthening access to quality antenatal care and global infection control efforts is crucial for reducing perinatal morbidity and mortality. However, as this was a cross-sectional study, the results show association and not causation. Further research is recommended.

Data availability statement

Data are available upon reasonable request. All data relevant to the study are included in the article or uploaded as supplementary information.