Maternal Feeding Practices of Children One to Three Years in Collins Chabane Municipality of South Africa

ABSTRACT

The aim was to assess maternal feeding practices of children one to three years. A descriptive observational design was employed. The sample consisted of mothers-child dyads. A validated structured questionnaire was used. Data was analyzed using SPSS version 26.0. The nutrition status of the children at birth indicated 11.6% underweight as compared to the time of the study (7.2%), 7.9% were stunted increased to 38.0%, while wasting decreased from 11.4%–2.4%. Early cessation of breastfeeding and inappropriate complementary feeding practices were the factors influencing growth. The prevalence of underweight and wasting were low while stunting and overweight were high.

KEYWORDS:

• Children one to three years
• growth
• maternal feeding practices
• South Africa

Introduction

Child growth is an important public health indicator for monitoring nutritional status within populations (Syed and Raafay 2010). Severe forms of malnutrition in childhood can lead to poor growth and development and have been recognized as one of the leading public health problems among children under five in developing countries (Amsalu and Tigabu 2008). The United Nations Children’s Fund (UNICEF 2019) reported protein energy malnutrition (PEM) to be the most common nutritional problem among preschool children. A similar observation has been reported by Kalimbira, Chilima, and Mtimuni (2006), who revealed that PEM in Malawi is one of the leading health problems. Chakraborty et al. (2006) also reported PEM to be higher amongst children aged one to 36 months in India.

Globally, 23% of children under five are stunted, 14% are underweight and in developing countries, 7% are wasted (World Health Organization 2016). Some children suffer from more than one form of malnutrition, such as stunting and overweight or stunting and wasting. There are currently no joint global or regional estimates for these combined conditions (UNICEF, WHO and World Bank 2020). The prevalence of malnutrition is still high in Africa with an estimated 30% of children being underweight and about 28% of children under five stunted (Kruger et al. 2012). The global trends are progressively improving: child stunting and underweight have reduced over the past two decades and the drop in poverty has been the key driver in this reduction (Ponce et al. 2018).

The relationship between maternal autonomy and child growth was assessed in a cross-sectional study of 422 mothers and their youngest children aged six to 24 months in the Bawku West District of Ghana (Saaka 2020). The study reported that healthcare autonomy better predicted child growth and dietary intake. Another qualitative study in West Africa explored socio-cultural practices and their influence on the feeding practices of 32 mothers and their children in Grand Popo, Benin (Lokossou et al. 2021). This study reported that mothers and children had monotonous diets with high consumption of vegetables and maize-based meals. Furthermore, food taboos, particularly during pregnancy, were revealed and those cultural beliefs were still followed by some mothers, and food rich in nutrients was pushed aside. Another study explored the association between mothers’ education and the World Health Organization’s (WHO) eight Infant and Young Child Feeding (IYCF) core indicators in six countries in South Asia (Tariqujjaman et al. 2022). Their findings demonstrated a strong positive association between a mother’s education and IYCF indicators. A review in India (Sangra et al. 2017) revealed that the common enablers of IYCF were higher maternal socioeconomic status and more frequent antenatal care visits of more than three. Common barriers to IYCF practices were low socioeconomic status and less frequent antenatal care visits. The review showed that the factors associated with IYCF practices in India are largely modifiable and multi-factorial. These studies attest to the need to ensure that mothers/caregivers have adequate knowledge of child nutrition.

In the Limpopo province of South Africa, 23.4% of children aged one to three years were found to be stunted, 11% underweight and 5% wasted in Capricorn District (Mamabolo et al. 2005). Modjadji and Madiba (2019) conducted a study in the same area in the Capricorn District and observed that 22% of children were stunted and 27% were underweight, while 27.4% of the mothers were overweight and 42.3% were obese. Mushaphi et al. (2008) showed that stunting (18,9%), underweight (7%) and wasting (7%) were present in Vhembe District of Limpopo province. Phooko-Rabodiba et al. (2019) similarly observed that there was a high (39.6%) prevalence of stunting among children under the age of 60 months, a medium prevalence of underweight, and a low prevalence of wasting in all children in Sekhukhune District of Limpopo province. They furthermore reported that most mothers/caregivers were overweight or obese. In terms of the socioeconomic status of the population, Sekhukhune is ranked lower than Vhembe and Capricorn Districts.

The South African Health Demographic Survey (SADHS) 2016 revealed that just over one-quarter (27%) of children under five were stunted nationally, 3% were wasted, while 6% were underweight and 13% were overweight (National Department of Health (NDoH), Statistics South Africa (Stats SA), South African Medical Research Council (SAMRC), and ICF, 2019). While the Limpopo stunting prevalence was 22% (NDOH, Stats SA, SAMRC, ICF 2019). Modjadji, Molokwane, and Ukegbu (2020) observed the prevalence of stunting (29%), underweight (13%) and thinness (6%) in the Northwest province of South Africa. Thus, the rates of malnutrition in Limpopo are comparable with other provinces and national levels. Undernourished children have lowered resistance to infection, poor wound healing, and poor clinical outcomes from disease or increased morbidity and mortality (Ibrahim et al. 2017). Malnourished children are more likely to die from diarrheal diseases and respiratory infections. On the other hand, undernutrition and micronutrient deficiencies have been associated with poor school performance and poor work capacity during adolescence and adulthood (Hoddinott et al. 2013). Chakona (2020) conducted a study to determine maternal dietary diversity, breastfeeding and IYCF practices in five rural communities in South Africa. They observed that social circumstances including lack of income, dependence on food purchasing, young mothers’ feelings regarding breastfeeding and cultural beliefs were the major drivers of mothers’ eating habits, breastfeeding behavior and IYCF practices. Another study was conducted to describe the infant feeding practices of South African women living in Soweto and to understand from the mothers’ perspective what influences feeding practices (Wrottesley et al. 2021). They found that feeding practices were influenced by mothers’ beliefs that what babies eat is important for their health. In addition, family matriarchs were highly influential in mothers’ feeding practices, however, their advice often contradicted that of health professionals. Modjadji and Madiba (2019) observed that having a younger mother and access to water and a refrigerator were protective against being underweight in Limpopo province. Furthermore, the nutrition knowledge and feeding practices of the mothers/caregivers were not satisfactory, as revealed in a study conducted in Limpopo province of South Africa (Motebejana, Nesamvuni, and Mbhenyane 2022).

A child’s ability to achieve standards in growth is determined by the adequacy of dietary intake which depends on infant and young child feeding practices, care practices and food security as well as exposure to disease (UNICEF 2011). Improving infant and young child feeding practices in children is critical to improved nutrition, health and development of children (Sangra et al. 2017). Adequate nutrition that includes exclusive breastfeeding for six months, appropriate complementary food and continued breastfeeding for at least two years and beyond are essential to improve the nutritional status of children and thus prevent child deaths((Motebejana, Nesamvuni, and Mbhenyane 2022). Infant and young child feeding practices affect the nutritional status of children and directly impact child survival according to Korday, Sharma, and Malik (2018). Sinhababu et al. (2010) found the proportion of appropriate feeding as per integrated management of neonatal and childhood illness protocol and WHO to be significantly low at 8.7% amongst children aged 12 to 23 months in India. Inappropriate feeding practices have been reported to be strongly associated with poor nutritional status in South Africa (Kleynhans, MacIntyre, and Albertse 2006). Furthermore, early introduction of solid feeds is associated with poor growth and risk of obesity, infections and allergies (Mamabolo et al. 2005).

Education for mothers and caregivers is essential to improve maternal knowledge and feeding practices. Mothers’ or caregivers’ feeding practices are influenced by knowledge, perception, cultural practices and attitudes according to the associated malnutrition found in rural Warda, Ethiopia with faulty feeding practices. They found that the intake of proteins and energy amongst children one to three years was 81% and 57% respectively.

Children one to three years of age are prime targets for intensified nutritional interventions (Labadarios et al. 2001). Furthermore, mothers/caregivers are targets for nutrition promotion and facilitation of healthy feeding practices. No statistics are indicating the prevalence of wasting and stunting in rural communities in Collins Chabane Municipality of South Africa. It is therefore important to investigate the current nutritional status of children one to three years in this rural region of the Limpopo Province. The study aimed to assess maternal feeding practices and nutritional status of children one to three years in Collins Chabane Municipality of Vhembe District, Limpopo Province in South Africa.

Methodology

Study design

The study employed a descriptive observational quantitative design. The study describes the growth and feeding practices of children one to three years old and the factors associated with growth and feeding practices. Furthermore, the relationship between growth patterns and feeding practices was determined.

Population and study area

The study population was mothers with children aged one to three years who use selected public primary health care clinics in Collins Chabane Municipality. At the time of the study, the area was served by 17 clinics and one health center serving 443 798 people of which 35.5% are under 16 years of age (Statistics South Africa 2011, 2023). The clinics were grouped in three clusters according to their geographic locations which are east, west and center. The health center was automatically included as the area consisted of one health center. Two clinics from each group were selected using the quota sampling method. A total of six clinics and one health center were then included in the study. The study was conducted in 2017.

Sampling design and procedure

The clinics were used as a sampling frame. Children were sampled using convenience sampling ensuring that all age groups are represented. All caregivers who gave concern for their children to be included in the study were included in the study. The exact number of children under three years attending the primary health care clinics at Collins Chabane Municipality was not readily available at the time due to demarcation politics that were going on in the Vhembe District municipality. A sample size of more than 100 subjects was targeted to allow for comparisons and statistical analysis (Anderson, Kelley, and Maxwell 2017; Maxwell, Kelley, and Rausch 2008). Thus, a total of 200 child-mother/caregiver dyads were envisaged in the study, with 50 dyads from the four age groups: 12 to 17 months, 18 to 23 months, 24 to 29 months and 30 to 36 months.

Data collection and procedure

Data was collected by the principal investigator who is a registered dietitian. Mothers/caregivers with children between one to three years were recruited from the Well Baby clinic and requested to participate in the study. Children and their mothers who satisfied the inclusion criteria of one to three years were selected. The procedure was explained by the researchers in Xitsonga language and caregivers who agreed to participate signed a consent form. Information on feeding practices was collected by using a food frequency questionnaire and a diet history. The anthropometric data measured and interviews were all done and recorded on the same day by the principal investigator. A structured interviewer-led questionnaire was adapted from that used in the preceding study (Mushaphi et al. 2008) and was translated into Xitsonga (the local language). The questionnaire had four sections which consisted of information on anthropometry and medical data, demographic data, and feeding practices and food frequency. General questions and a food frequency questionnaire were used to estimate children’s feeding practices and disease status. The translated questionnaire was tested on five mothers/caregivers at a hospital where the principal investigator is based.

Each clinic was visited three times, first for introductions and twice for data collection. Data was collected over four weeks. The clinic management allocated one of the consultation rooms for use. The principal investigator interviewed the participants using the local language. The whole process from recruitment, consenting process, measurements and interview took approximately one and a half hours.

Anthropometric measurements

The birth weight of the children as recorded on the road to health charts was used. The Road to Health Chart in South Africa is a booklet that is created for children at birth to record medical history, health, immunizations, growth and development from birth to age five (Win and Mlambo 2020). Weight, height and MUAC for children were measured following the standard protocol of the International Society for the Advancement of Kinanthropometry (ISAK) (Silva and Vieira 2020). The weight and height of the mothers/caregivers were also measured using the standard protocols. The weight and height of the caregivers were used to calculate body mass index (BMI) and to estimate the mothers’ nutritional status. The mothers’ nutritional status was measured to control for bias introduced by confounders/modifiers.

Data analysis

Data was analyzed using the Statistical Package for Social Sciences (SPSS) for Windows version 25.0. Anthropometric z-scores were analyzed using WHO Anthro 1.0.4. Descriptive and inferential statistics were utilized to analyze data according to age groupings. Weight-for-age, height-for-age and weight-for-height were analyzed using z-scores (31) (WHO 2006). A child is malnourished if their z-score is < −2, while severe malnutrition is indicated at a z-score of < −3 (28) (WHO 2006). Arm circumferences were analyzed using the National Centre for Health Statistics (NCHS) tables (32).

Descriptive statistics that were used are the mean standard deviation, range and percentiles. The chi-square was used to compare feeding practices, anthropometric measurements and socio-economic factors. The t-test was used to determine the significance of the mean weight, length, and arm circumference and the difference between the mean standard references by age group. The results are significant if the p-value is < 0.05.

Ethical considerations

Approval was sought from the higher degrees and ethics committee of the University of Venda. Permission to use the clinics was requested from the Department of Health at the Provincial Office and from the Primary Healthcare Manager. Consent forms were signed by caregivers to permit participation of themselves and their children. Data was treated with confidentiality and a code was assigned to the mother-child dyad. Children who were found to be at risk of malnutrition were referred for treatment and/or nutrition counseling.

Results

The study comprised 141 children and 141 mothers/caregivers from six local clinics and one health center. This was 70.5% of the envisaged sample. The age group above two years did not frequent the primary health facilities. The age of the children ranged from 12 to 36 months with the majority of the children (65.2%) below the age of 24 months while 34.6% were between the ages of 24 to 36 months. Regarding children’s gender, more than half (55.7%) were male and 44.3% were female. See Table 1.

Table 1. Sociodemographic characteristics of the participants (n = 141).

Variables	Number (n = 141)	Per cent (%)
Age of children
12–17 months	46	32.6
18–23 months	46	32.6
24–29 months	21	14.8
30–36 months	28	19.8
Age of the mothers(years)
13–17	3	2.1
18–22	40	28.4
23–27	30	21.3
28–31	26	18.4
32–35	24	17.0
35 and above	18	12.8
Marital status of the mother
Single	38	27
Married	79	56
Divorced	9	6.4
Widow	5	3.5
Live with partner	10	7.1
Level of education of the mother
Never attended	13	9.2
Grade 1–4	16	11.3
Grade 5–7	16	11.3
Grade 8–10	33	23.4
Grade 11–12	59	42.0
Tertiary	4	2.8
Breadwinner of the child’s household
Father	27	19.1
Mother	56	40.0
Father and mother	31	22 .0
Grandparents	22	16.0
Mother and grandparents	4	2.8
Other family members and friends	1	0.7
Employment of the breadwinner
Social grants	63	45.0
Government worker	23	16.2
Self-employed	12	8.5
Private	21	15.0
Domestic worker	22	15.6
Child’s household monthly income ZAR (in Rands and US dollars)
Less than 499 (27 USD)	30	21.3
500–999 (27 to 55 USD)	57	40.4
1000–1499 (55 to 82 USD)	31	22.0
1500– 1999 (82 to 110 USD)	7	5.0
2000– 2499 (110 to 137 USD)	3	2.1
2500 and above (>137 USD)	10	7.1
Does not know	3	2.1

USD = United States Dollar; ZAR = South African Rands.

Most mothers (85.1%) were of child-bearing age as they were between the ages of 18 and 35 years. More than half of the mothers/caregivers (56%) were married, and 68.2% were literate as they had secondary education (grades 8–12). Most (81.1%) children were supported by their parents, with 45% of the households dependent on social grants whilst 16.2% depended on family members working in government institutions. Nearly two-thirds of households (61.7%) had an income of less than 1000.00 ZAR (55 USD) (see Table 1).

Health profile of the children

About one-third (36.2%) of the children were reported to suffer from respiratory infections, diarrhea (13.5%), diarrhea combined with respiratory infections (22.7%), diarrhea and vomiting (22.7%), and vomiting (2.8%) while 2.1% did not have any condition.

Child growth patterns

Nutritional status of the children at birth indicated that more than half of the children were born with normal weight (57.6%), mild underweight (18.7%), stunting (7.9%), underweight (5.8%) and severely underweight (5.8%), while children who were overweight were 12.1%.

Weight-for-age at the time of study indicates that only 1.4% of the children were severely underweight. This shows that underweight has improved from 5.8% at birth to 1.4% as the children were growing. Although 44.3% of the children had normal weight at the time of study, 26.6% of the children showed a possible faltering growth problem.

Regarding stunting prevalence, detailed analysis revealed that 2.9% were severely stunted and 5.0% were stunted at birth. At the time of the study, severe stunting increased from 2.9% at birth to 20.1% and stunting increased from 5.0%–17.9%.

The prevalence of wasting was 5.7% and severe wasting (5.7%) at birth. The possible risk of overweight was 12.9% and overweight 9.3%. Analysis revealed that at the time of the study, the prevalence of severe wasting was 1.5% and wasting 0.7%. The risk of being overweight was 26.6%, overweight 15.8% and obesity 6.5%. Wasting decreased by 9.2%, obesity increased by 6.4%. overweight increased by 6.5%, and the risk of overweight increased by 6.5% from birth to the time of the study. Mid-upper arm circumference of the children indicated that most of the children were normal. See Table 2.

Table 2. An illustration of the nutritional status of children one to three years at birth and during the study (n = 141).

		At birth		At the time of the study
z-score classification	Interpretation	n	%	n	%
Z-score classification of weight for age (WAZ)
>+1SD	Overweight	17	12.2	37	26.6
>-1SD; +1SD	Normal WAZ	80	57.6	62	44.6
−2SD; ≤ −1SD	Mild underweight	26	18.7	30	21.6
≥-3SD; ≤-2SD	Underweight	8	5.8	8	5.8
<-3SD	Severely underweight	8	5.8	2	1.4
Z-score classification of height for age (HAZ)
>-2SD	Normal HAZ	129	92.1	86	62.0
≥3SD; ≤-2SD	Stunted	7	5.0	25	17.9
<-3SD	Very stunted	4	2.9	28	20.1
Z-score classification of weight for height (WHZ)
>3SD	Obese	0	0	9	6.5
≥+2SD; <3SD	Overweight	13	9.3	22	15.8
≥+1SD; <2SD	At risk of wasting	18	12.9	37	26.6
≥-2SD; <+1SD	Normal	93	66.4	68	48.9
≥-3SD; <-2SD	Wasting	8	5.7	1	0.7
<-3SD	Severe wasting	8	5.7	2	1.5
The mid-upper arm circumference (MUAC)
0–11.5 cm	Severe acute malnutrition	–	–	1	0.7
11.5–12.5 cm	Moderate malnutrition	–	–	1	0.7
>12.5 cm	Normal	–	–	138	98.6

WAZ: weight for age; HAZ: height for age; WHZ: weight for height; MUAC: mid-upper arm circumference; n = sample size; SD: standard deviation.

Feeding practices of mothers/caregivers

Most of the children were fed by their mothers (58.2%) with 31.9% self-feeding. Regarding the duration of child breastfeeding, 36.2% of the children were still breastfed at the time of the study with 3.2% of mothers breastfeeding beyond 18 months. Most (57.5%) mothers breastfed their children between six to 18 months (see Figure 1).

Figure 1. shows the duration of breastfeeding among mothers in Vhembe District.

Dietary patterns of children

Concerning milk consumption, most of the children (71.6%) were not fed any milk whilst few were given pasteurized cow’s milk (15.6%). Snacks fed to babies were also assessed and findings revealed that dried chips, biscuits, sweets and sweetened drinks were most fed to the children (45.4%). Concerning the type of salt used in the household, most of the households (59%) were using iodated salt and 33.3% non-iodated salt.

Food frequency was assessed by asking mothers/caregivers how often they gave certain food to their children. As shown in Table 3, food groups mostly consumed by the children were cereals, fruits and vegetables. Detailed analysis revealed that the top ten foods mothers/caregivers gave to the children daily were stiff porridge, soft porridge, brown bread, peanut butter, fat cakes, sugar-sweetened beverages and snacks, margarines, fruit and fruit juices. Other foods given at least once or twice a week included spinach, indigenous vegetables mixed with peanuts, eggs, potatoes, cabbage, beans, fish, indigenous vegetables, soya mince soup, rice, fruits, chicken and yogurt. Traditional foods were amongst the least foods fed to the children. This pattern shows that maize is utilized daily while the accompaniment is variable. The variety seems limited and influences the children’s eating habits.

Table 3. The food consumption of children one to three years at birth and during the study in Vhembe District (n = 141).

Food item	Daily %	Weekly %	Occasionally %	Never %
Cereals
Stiff porridge	91.3	5	0	3.6
Soft porridge	81.2	8.7	0.7	8.7
Breakfast cereals	2.2	0.7	1.4	95.7
Samp	0.7	12.3	10.1	76.8
Rice	0.7	42.7	12.3	44.2
Sweet potatoes	0.7	10.1	8.0	81.2
Potatoes	0	52.9	10.9	36.2
Baby foods	1.4	5.7	2.9	89.9
Pasta & macaroni	3.6	6.5	2.2	87.7
Brown bread	51.4	23.9	2.2	22.5
White bread	4.3	3.6	2.9	89.1
Dairy
Yoghurt	2.2	21.7	24.6	51.4
Milk	8.0	8.6	5.8	77.5
Cheese	0	3.6	2.2	94.2
Ice cream	2.2	7.9	88.4	0.7
Animal foods
Chicken	0	37	20.3	42.8
Chicken livers	0.7	6.5	3.6	89.1
Mopani worms	0	3.6	3.6	89.1
Termites	0	1.4	1.4	97.1
Locusts	0.7	6.5	2.2	90.6
Fish	2.2	49.3	9.4	39.1
Red meat	0	18.1	22.5	58.7
Minced meat	5.8	4.4	1.4	90.6
Polony	1.4	0.7	0.7	97.1
Eggs	0.7	53.9	15.1	29.5
Chicken feet	0.7	10.1	20.9	67.6
Legumes
Beans	1.4	49.3	10.1	39.1
Soya mince soup	4.3	42.7	5.8	47.1
Peanuts	5.1	23.2	3.6	68.1
Fruits and vegetables
Fruits (variety based on season)	17.4	42.7	7.2	32.6
Fruit juices	16.7	15.9	8.7	58.7
Peanut butter	24.6	20.3	2.9	52.2
Pumpkin	0	15.9	2.2	81.9
Cabbage	1.4	50	10.1	38.4
Spinach	0.7	56.5	5.8	37
Carrots	0	12.3	4.3	83.3
Beetroot	1.4	18.1	1.4	79
Indigenous vegetable	5.8	48.6	1.4	44.25
Indigenous vegetables mixed with peanuts	6.5	54.3	1.4	37.7
Other food items
Jam	2.2	7.2	1.4	89.1
Butter	0	1.4	2.2	96.4
Margarine	18.1	16.7	2,9	62.3
Cold drinks/sweets/cheese snacks/biscuits	19.6	32.6	6.5	41.3
Fat cakes	23.9	26.8	4.3	44.9
Indigenous food
Tihove (mixed beans and nuts dish)	2.9	2.8	12.3	81.9
Tshopi (pumpkin and maize dish)	0	3.6	7.2	89.1
Xingwimbi (pumpkin and rough maize dish)	0	0.7	0.7	97.1

Relationship among socioeconomic factors, feeding practices and nutritional status of children

Table 4 below indicates wasting was positively influenced by the age of the mother, toilet facility and the source of fuel for cooking, and negatively influenced by the maternal level of education (p < .05). Stunting was not significantly associated with any socioeconomic variable in this study (p > .05). Underweight was positively influenced by the feeder of the child and the availability of storage facilities in the households (p < .05). The birth weight of the child has a positive correlation with the child’s household source of drinking water and the type of toilet facility used by the household. There was a positive influence of the household income and the availability of food in the household on MUAC (p < .05).

Table 4. Shows the association between socioeconomic factors of the nutritional status of children (Pearson’s correlation).

Variables	Wasting (WHZ) r(p)	Stunting (HAZ) r(p)	Underweight (WAZ) r(p)	Body mass index (BAZ) r(p)	MUAC2 r(p)
The feeder of the child	-.102 (0.234)	-155 (0.068)	-.1.88 (0.027) *	-.096 (0.263)	-.029 (0.741)
Maternal age	-173 (0.041) *	-.075 (0.379)	-.019 (0.823)	.150 (0.077)	-.086 (0.325)
Marital status of the mother	-.059 (0.049) *	-.083 (0.334)	-.113 (0.184)	-.093 (0.276)	-155 (0.072)
Maternal level of education	-.004 (0.961)	-.054 (0.525)	-.024 (0.778)	-.043 (0.611)	-.004 (=.965)
Child’s source of income	-058 (0.496)	-.055 (0.518)	-.074 (0.389)	-.047 (0.584)	-.124 (0.150)
Mother’s employment status	-.137 (0.107)	-.038 (0.660)	-.096 (0.259)	-.163 (0.056)	-.114(0.188)
Household income	-.092 (0.282)	-.059 (0.251)	-.122 (0.152)	-.139 (0.107)	-.534 (0.001) **
Household size	-062 (0.467)	-.049 (0.563)	-.079 (0.353)	-.076 (0.374)	-.058 (0.502)
Type of livestock reared in the household	-.042 (0.623)	-.004 (0.961)	-.021 (0.805	-.046 (0.588)	-.076 (0.377)
Households own land for food production	-.096 (0.263)	-.062 (0.468)	-.098 (0.249)	-.088 (0.305)	-.023 (0.788)
Household source of drinking water	-.169 (0.046) *	-.068 (0.425)	-.040 (0.637)	-.183 (0.031) *	-.074 (0.390)
Household toilet facility	-.215 (0.011) *	-.047 (0.586)	-.039 (0.648)	-.175 (0.039) *	=.081 (0.347)
Household used electricity	-.079 (0.352)	-.079 (0.355)	-.065 (0.447)	-.028 (0.745)	-.065 (0.445)
Household-owned cold storage facility	-.069 (0.423)	-.159 (0.061)	-.167 (0.049) *	-.067 (0.433)	-.116 (0.178)
Source of receiving public information	-.153 (0.072)	-.041 (0.633)	-.106 (0.214)	-.155 (0.069)	-.167 (0.053)
Source of fuel for cooking	-.177(0.037) *	-.075 (0.383)	-.015 (0.861)	-.142 (0.097)	-.010 (0.907)
The duration the child was breastfed	-.055 (0.520)	-.071 (0.406)	-.103 (0.228)	-. 064 (0.456)	-.056 (0.514)
Feeding the child with other milk	-.036 (0.674)	-.025 (0.768)	-.036 (0.675)	-.043 (0.613)	-.076 (0.378)
Type of milk used to feed the child	-.082 (0.336)	-.001 (0.989)	-.038 (0.658)	-.063 (0.459)	-.121 (0.162)
The frequency of feeding a child with milk per day	-.077 (0.367)	-.106 (0.214)	-.120 (0.158)	-.066 (0.441)	-.109 (0.208)
Availability of food at home	-0.83 (0.334)	-.008 (0.923)	-.089 (0.300)	-.160 (0.055)	-.220 (0.009) **
Lack of child food/food insecurity	-.125 (0.143)	-.041 (0.634)	-.107 (0.211)	-.131 (0.123)	-.042 (0.623)
Households ever run out of money to buy food	-.090 (0.294)	-.020 (0.814)	-.034 (0.691)	-.084 (0.323)	-.110 (0.203)

Bold* depicts significance at 0.05 and Bold** depicts significance at 0.005. WHZ: weight-for-height; HAZ: height-for-age.

WAZ: weight-for-age; BAZ: body mass index for age

Wasting was negatively influenced by feeding soft porridge and fruit juice, while positively influenced by stiff porridge and margarine. Stunting was negatively influenced by feeding bread and fruits. Underweight was negatively influenced by feeding bread, fruits and cold drinks, sweets and biscuits. Birth weight z-scores were negatively influenced by feeding soft porridge and fat cakes and positively influenced by stiff porridge, potatoes and margarine (Table 5).

Table 5. The relationship between growth and feeding practices among children one to three years in Limpopo, South Africa (Pearson’s correlation, p < .05).

Variables	Wasting (WHZ) r(p)	Stunting (HAZ) r(p)	Underweight (WAZ) r(p)	Body mass index (BAZ) r(p)
Stiff Porridge	-.673 (0.036)	-.130 (0.130)	-.128 (0.135)	-.606 (0.044)
Soft porridge	-.958 (0.0025)	-.748 (0.028)	-.965 (0.004)	-.708 (0.032)
Potatoes	-.627 (0.042)	-.145 (0.091)	-.133 (0.120)	-.862 (0.015)
Bread	-.385 (0.075)	-.958 (0.005)	-.568 (0.049)	-.288 (0.091)
Fruits	-.428 (0.068)	-.741 (0.029)	-.956 (0.005)	-.524 (0.055)
Fruit juices	-.768 (0.025)	-.466 (0.063)	-.976 (0.003)	-.534 (0.054)
Peanut butter	-.151 (0.123)	-.152 (0.076)	-.425 (0.069)	-.127 (0.138)
Margarine	-.811 (0.021)	-.710 (0.032)	-.930 (0.008)	-.897 (0.011)
Cold drinks/sweets/cheese snacks/biscuits	-.402 (0.072)	-.831 (0.018)	-.863 (0.015)	-.497 (0.059)
Fat cakes	-.518 (0.056)	-.406 (0.072)	-.541(0.053)	-.774 (0.025)

bold depicts significance; WHZ: weight-for-height; HAZ: height-for-age; WAZ: weight-for- age; BAZ: body mass index for age.

Multiple regression analysis using four models showed that age was the only variable associated with feeding practices. The older the mother, the better the feeding practices. See Table 6 below.

Table 6. Multiple regression using maternal age and BMI and MUAC and feeding practices.

Coefficients^a
				Standardized Coefficients			95.0% Confidence Interval for B
Model				Beta	t	Sig.	Lower Bound	Upper Bound
1	(Constant)	21,811	6,598		3,306	0,001	8,764	34,859
	age	0,197	0,088	0,202	2,234	0,027	0,023	0,372
	MUAC	−0,782	0,453	−0,159	−1,726	0,087	−1,679	0,114
	maBMI	0,206	0,126	0,140	1,632	0,105	−0,044	0,456
2	(Constant)	24,503	6,427		3,813	0,000	11,794	37,211
	Age	0,180	0,088	0,184	2,043	0,043	0,006	0,355
	MUAC	−0,611	0,444	−0,124	−1,376	0,171	−1,488	0,267
3	(Constant)	16,053	1,905		8,429	0,000	12,287	19,819
	Age	0,136	0,083	0,139	1,651	0,101	−0,027	0,300
4	(Constant)	19,029	0,619		30,750	0,000	17,805	20,252

MUAC: medium upper arm circumference; maBMI: maternal body mass index.

Health status of the mothers

Regarding the BMI of mothers, 62.5% had normal BMI (18.6–25 kg/m²), 18.1% were overweight with BMI above 26–29 kg/m², 13.3% were obese (30–47 kg/m²) while few mothers (5.7%) were underweight (BMI: 16–18 kg/m²).

Discussion

The study revealed the prevalence of stunting with low levels of wasting and underweight and an increase in overweight rates. The findings showed that mothers/caregivers fed children a high starchy diet with a limited variety of foods. Associations between growth patterns and feeding practices were weak.

Socio-demographic characteristics

Findings indicated that most children in this study were from households with low income as social grants and income from domestic work contributed as the main source of support for most families. The findings are consistent with preceding findings which reported that 66% of South African children live in households that have less than 1200 ZAR (66 USD) per month (Labadarios et al. 2005; Phooko-Rabodiba et al. 2019). Low income and unstimulating household environments contribute to deficits in children’s development and health, and productivity in adulthood (Black et al. 2013). This can be explained by a limited intake of food by the poorest due to shortage and scarcity.

Regarding the BMI of mothers, about two-thirds of the mothers had normal weight and there were low rates of overweight and obesity. This study contradicts the findings of a previous study conducted in both the urban and rural southern Free State of South Africa which reported that one-third of the rural mothers/caregivers were normal/underweight with the majority being overweight or obese (Tydeman-Edwards, van Rooyen, and Walsh 2018). Kaldenbach et al., found that a double burden of malnutrition was prevalent in KwaZulu-Natal, with most mothers being overweight or obese, and many children developing stunting in the early months of life. The increasing rates of overweight and obesity in South Africa pose serious challenges for non-communicable diseases. However, there seem to be regional variations in the prevalence. One of the weaknesses of this study was the lack of comparison of mothers’ BMI and children’s z-scores. The double burden of malnutrition in the same household was thus not determined. Akokuwebe and Idemudia (2022) conducted a cross-sectional survey in Nigeria and South Africa. Their multilevel analysis revealed urban-rural variations of body weights and individual-level factors among women of childbearing age.

Health profile of the children

Despite the high attendance rate of six-week immunization services in South African facilities, potential missed opportunities for early infant diagnosis are still very high (Woldesenbet et al. 2015). Inadequate communication between antenatal, delivery, and postnatal facilities and poor information systems have been reported as important barriers to the continuity of postnatal services in developing countries (Horwood et al. 2010; Reithinger et al. 2007).

Health problems experienced by the children during the period of the survey included respiratory infections such as coughing, diarrhea and vomiting. The Committee on Morbidity and Mortality in children under five years reported diarrhea, lower respiratory tract infections, neonatal conditions with HIV/AIDS and malnutrition contributing as primary and underlying major causes of child mortality (Victora et al. 2008).

Growth patterns of children

The z-scores indicated that one-tenth of children were underweight at birth compared to less than one-tenth at the time of study. This might imply that underweight decreases with age as some of the underweight children can catch up with growth. Similar findings have been reported by Mamabolo et al., (2005) where it was revealed that the prevalence of underweight children under three years is 9.0% in rural Limpopo province. The SADHS of 2016 also reported a low underweight of 6% (NDOH, Stats SA, SAMRC AND ICF, 2016). Child feeders and the availability of storage facilities in the households were the major determinants of underweight in this study.

The present finding revealed that less than one-tenth of the children were stunted at birth and this increased five times during the time of study. The implication is that an increased number of children might experience impaired neurocognitive development, perform poorly at school and reduced productivity in later life (World Health Organization 2014). This prevalence was lower when compared to a study conducted earlier which reported the prevalence of stunting of almost half among children one to three years in the same geographic area (Mamabolo et al. 2005). The findings in this study are still higher than the stunting trends reported in the 2016 SADHS in Limpopo province of 22% (NDOH, Stats SA, SAMRC AND ICF 2016). Furthermore, the prevalence is also higher than the national stunting rate of 27.0% for children under five years (Statistics South Africa 2016).

The prevalence of wasting was one-tenth at birth and decreased to 2% at the time of the study. This infers that fewer children were too thin for their height. Wasted children have weakened immunity, are susceptible to long-term developmental delays, and face an increased risk of death, especially when wasting is severe (UNICEF, WHO and World Bank 2020). The study findings are in line with the national prevalence of 3% (Statistics South Africa 2016). A consistent association in low-income countries has been demonstrated between maternal age, maternal level of education, sanitation and wasting (Das et al. 2021; Melese 2014; Makoka and Masibo 2015). This study showed no associations between wasting and maternal age, level of education and sanitation.

The prevalence of overweight at the time of the study was just above 20% from just below 10% at birth. Thus, the prevalence of overweight increases as they grow older. Jebeile et al. (2022) assert that the development and continuance of obesity are explained by a bio-socioecological framework, whereby biological predisposition, socioeconomic, and environmental factors interact together to promote the deposition and proliferation of adipose tissue. The prevalence of overweight in this current study is slightly lower when compared to the 24% prevalence of overweight across under-five-year-old children in Africa (World Health Organization 2022).

MUAC was less than 2%, signifying the rate of acute malnutrition was low among the children. These results are in contrast with findings in some low-income settings in Africa that show a higher prevalence of MUAC in children (Engebretsen et al. 2008; Kinyoki et al. 2015). There was a positive influence of the household income and the availability of food in the household on MUAC (p < .05) in this study. Income results in higher purchasing power and a household’s ability to afford diverse foods (Nascimento et al. 2021).

Breastfeeding and complementary feeding practices

There seems to be early cessation of breastfeeding amongst the mothers in this study group as results showed almost all the children were initially breastfed but weaned off by six to 18 months. The SADHS 2016 reported the median breastfeeding period as 13.3 months in rural areas of South Africa (NDOH, Stats SA, SAMRC AND ICF 2016). Thus, this study was comparable to the national trends. Infant feeding practices among mothers seem not to follow the WHO recommendation or IYCF of continued breastfeeding for two years (UNICEF 2011). Early cessation of breastfeeding in favor of commercial breastmilk substitutes, the introduction of liquids such as juices and poorly timed complementary feeds of poor quality seem to be common (Cai, Wardlaw, and Brown 2012). González-Cossío et al. (2006) study found the breastfeeding rates of mothers in Mexico to be at 92.3% at initiation but they declined at one year to less than 40% of the children still breastfed. Breastmilk is a natural resource that has a major impact on a child’s health, growth and development and is recommended for at least two years of a child’s life. Breastmilk provides half or more of a child’s energy needs between six and 12 months and one-third of energy needs between 12 to 24 months (Dewey 2005). Breastmilk is also a critical source of energy and nutrients during illness and reduces mortality among children who are malnourished (World Health Ogarnization 2013).

Linear growth retardation which leads to stunting often begins in utero but is most marked during the vulnerable period of complementary feeding, the transition from a diet of breastmilk to family food (Victora et al. 2008). Continued breastfeeding and complementary feeds usually meet protein requirements but energy intakes may be low, and diets are almost low in critical micronutrients (Dhami et al. 2021; Victora et al. 2008). Information on child feeding practices in this study revealed that a low variety of locally available foods are fed to children but included mostly starchy foods. Foods mostly used were porridge, bread, protein foods like chicken, eggs, fish and soya mince soup, fruits and vegetables, especially traditional green leafy vegetables, cabbage and spinach.

More than half of the children were no longer breastfed and other non-human milk was not commonly fed to the children. In this study, only just over half were less than two years of age and expected to be still breastfeeding. Children who are not breastfed should be given one to two cups of milk and one to two extra meals per day (WHO 2001). The implications of not giving children milk will be a lack of calcium and poor bone development. Milk also provides energy, protein, and vitamins A, B and D. These nutrients are required for bone and teeth development, growth, vision, immune system and heart function. Poor milk consumption by children could lead to growth retardation (WHO 2001). In addition, dried chips, sweets, biscuits and sweet drinks were the most common snacks fed to children. Similar findings were reported in another study which found that children are given milk until around the age of one year and thereafter milk falls off the children’s diet and sweets and crisps are commonly fed to children at three years (Mamabolo et al., 2005). Age of mother was the only variable shown to influence feeding practices. Dhami et al., (2019) showed that older mothers (≥25 years) in the Northern region of India were more likely to have children who met minimum adequate diet compared with younger mothers (<25 years). Quamme and Iversen (2022) showed no significant link between maternal age and stunting in Sub-Saharan Africa. In contrast, Nankinga et al. (2019) reported that maternal age of 35–49 had lower odds of stunting compared with children born to younger mothers aged 15–24 years (OR 0.69, 95% CI 0.56–0.86) in a study conducted in Ghana. Other studies in Sub-Saharan Africa have linked older age to better feeding practices potentially due to better nutrition knowledge, parenting experience and higher incomes (Adedokun and Yaya 2021; Masuke et al. 2021; Tesema et al. 2021).

This study was limited to children one to three years old attending the public health institutions in Collins Chabane Municipality because they were the accessible population. In addition, the sampling frame of clinics was probably a limitation. It is possible that some children above two years were not attending the Well Baby clinics as it often happens after the immunizations are completed. The authors also acknowledge that child growth is not only influenced by maternal feeding practices and the other confounders were not controlled in this study.

Conclusion

The prevalence of underweight, wasting and MUAC were low while stunting and overweight were high among children one to three years at the time of the study when compared to the provincial and national rates. The high stunting and overweight among this age group should be targeted for intervention. Information on child feeding practices indicated that mostly starchy foods with limited non-human milk, fruits and vegetables were fed to children. Inappropriate breastfeeding and complementary feeding practices were the major determinants of poor growth in children. This necessitates an urgent call for interventions to target mother/caregiver nutrition education and empowerment.

Acknowledgments

The authors thank all the mothers and children who participated in this study for their cooperation and support.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data is available from the corresponding author [email protected].

Additional information

Funding

Departmental funding from Human Nutrition at the University of Venda.