This analytical cross-sectional study sought to measure the prevalence of complementary feeding (CF)-related WHO/UNICEF IYCF indicators and child nutritional status, to identify factors associated with child nutritional status and to determine the association of the IYCF indicators with child nutritional status among children (6–23 months) in the Northern Region of Ghana, after adjusting for potential confounders. Bivariate and multivariable statistical analysis of the associations between the CF-related IYCF indicators and the nutritional status of children in northern Ghana were thus examined.
4.1. Prevalence of IYCF Indicators among Children (6–23 Months) in Northern Ghana
One of the main findings of this study was that, generally, the prevalence of the CF-related IYCF indicators estimated in the Northern Region of Ghana were relatively higher compared to the national and northern regional findings of the GDHS 2014 [
5]. The prevalence of stunting remained unchanged compared to the GDHS 2014 findings, but the levels of wasting and underweight showed increases, despite improvements in the prevalence of the CF-related WHO/UNICEF core IYCF indicators.
The IYCF indicators are nutrition-specific indicators recommended by the WHO, UNICEF and other maternal and child nutritional health experts as suitable, highly cost-effective, evidence-based and pragmatic parameters for monitoring and evaluating progress in child nutritional health in order to inform the development of appropriate interventions against the various forms of undernutrition [
4,
9]. The findings of this study showed some interesting similarities and disparities compared with the IYCF indicators reported in some recent studies conducted both in the Northern Region of Ghana and nationally. The Multiple Indicator Cluster Survey Six (MICS 6) survey, conducted in 2017–2018, reported a national prevalence of 79%, 41%, 23% and 12% respectively for TICF, MMF, MDD and MAD among children aged 6–23 months [
30]. The GDHS 2014, conducted in 2013, reported a prevalence of 45.2%, 17.9%, 14.1%, 47.7% and 55.9% for MMF, MDD, MAD, iron and vitamin A, respectively, among children (6–59 months) in the Northern Region [
5]. The indicators from our study were generally higher compared with the GDHS 2014 findings for the Northern Region. In a nutrition surveillance (NS) study, conducted in November 2013, of the three northern regions of Ghana, Saaka et al. [
14] also reported a prevalence of 48.8%, 58.2%, 34.8%, 27.8% and 15.7% for TICF, MMF, MDD, MAD and ACF respectively among children aged 6–59 months. The general improvements in IYCF indicators reported from our study could possibly reflect the various nutritional intervention strategies implemented in northern Ghana over the period between November 2013 and June 2018, when these studies were conducted.
4.2. Nutritional Status of Children (6–23 Months) in the Northern Region of Ghana
The prevalence of undernutrition in the Northern Region generally has not improved compared to that found in recent studies. For instance, compared to the estimated prevalence in the GDHS 2014 of stunting (33.1%), wasting (6.3%) and underweight (20%) among children (6–59 months) in the Northern Region of Ghana, stunting has remained the same, wasting has increased by a factor of about two, and underweight has also increased, by almost one-third [
5]. The Multiple Indicator Cluster Five (MICS 5) survey, conducted in 2011, reported 37.4%, 8.1%, 24.2% and 1.1% for stunting, wasting, underweight and overweight prevalence respectively in the Northern Region [
31]. MICS 6, conducted in 2017–18, reported 29%, 9% and 1% for stunting, wasting and overweight among children (6–59 months) in the Northern Region [
30]. Saaka et al. [
14] reported a prevalence of 20.5%, 11.5% and 21.1% for stunting, wasting and underweight respectively in a survey of 44 districts in the northern regions of Ghana conducted in November 2013 among about 1.3 million children under the age of five years. The disparities could be attributed to the age ranges of the children surveyed in these studies or the periods of the surveys, among other reasons. A disaggregation of the prevalence rates for children under two years of age (6–24 months) from the GDHS 2014 and the MICS 5 and 6 studies would be optimal for comparison with this study. Nonetheless, chronic undernutrition (LAZ), seems intractable in the Northern Region of Ghana. The majority of the stunted, wasted, underweight and overweight children were aged 12–23 months (
Table 3). This observation supports the hypothesis that the peak period for complementary feeding among growing children, with or without continued breastfeeding and comorbidities, is critical to optimal child growth [
32].
Compared to the GDHS 2014, MICS 5 and MICS 6, the prevalence of wasting (WAZ) in the Northern Region has increased [
5,
30,
31]. Acute undernutrition or wasting (WLZ), which reflects the effect of acute inadequate nourishment among children, is a more useful measure in emergency situations than stunting, which reflects an impairment of growth due to prolonged inadequate nutrient intake. The level of wasting estimated in this study (14.1%) is sufficient to be classified as serious (10–14%) based on the WHO cut-off values for public health significance [
2,
6]. This is probably due to the period of the survey (June), during which the onset of rain ushers in the planting season, which is known for its scarcity of food and increase in infectious diseases such as malaria.
Underweight (WAZ) and overweight (WLZ) status were both present in this study population, at 27% and 2.6% respectively. Falling within the high prevalence category according to the WHO public health significance cut-off (20–29%), the level of underweight is a phenomenon reflective of the overall nutritional health of children, which is usually premised on one or both of chronic growth impairment and acute growth impairment due to prolonged famine, acute hunger or frequent illness, among other causes [
2,
6]. Although the prevalence of overweight status among the children was relatively low, the emerging phenomenon of the double or multiple burdens of malnutrition in developing countries calls for efforts to stem the tide before it becomes too widespread in the Northern Region. However, generally, the prevalence of the IYCF indicators is relatively higher, this did not translate into a decrease in the prevalence of stunting or wasting status of the children in northern Ghana. Instead, the indicators of undernutrition were either similar to or worse than before. Appropriate, context-specific interventions are therefore needed in order to address this public health menace in northern Ghana.
4.3. Association between IYCF Indicators and Undernutrition (Stunting and Wasting)
Intake of iron-rich foods, child age group and maternal height were significant predictors (
p < 0.05) of stunting among the children studied in the Northern Region. Child gender, child age and the source of power for lighting in households were significant predictors (
p < 0.05) of wasting among the children. None of the IYCF indicators showed a significant association with stunting or wasting after adjusting for potential confounders in the models, except intake of iron-rich foods for stunting (
Table 6 and
Table 7). This apparent lack of significant correlation between the IYCF indicators and child nutritional status may also be due to insignificant variations in the child feeding practices of the study population generally. The study was conducted during the lean season and dietary diversity may be altered during this period because most rural households depend on their own production of subsistence crops, livestock and birds. However, some studies in SSA and SEA countries found strong relationships between some of the complementary feeding (CF)-related indicators and nutritional status, while others reported no significant associations [
25,
33,
34,
35]. The inconsistent findings about the relationship between the CF-related IYCF indicators and child anthropometric growth indicators (LAZ, WLZ and WAZ) may thus be context-specific and moderated by other factors or covariates [
4,
36,
37,
38].
Similar to this study, Saaka et al. [
14], in their quest to ascertain the relationship between the IYCF indicators and nutritional status of children in northern Ghana, found that none of the core CF-related indicators was a significant predictor of stunting (LAZ), except TICF. Interestingly, in this same study by Saaka et al. [
14], ACF, a composite food intake (CFI) metric estimated from TICF, MMF and MDD, was a significant predictor of wasting, but not stunting. Reinbott et al. posit that a composite child feeding index (CCFI), computed from a combination of the IYCF core indicators, is superior to the individual IYCF indicators in explaining stunting among children in DCs [
25]. The predictive utility of variously computed CCFIs from the individual IYCF indicators should be further investigated to validate or disprove the hypothesis that CCFIs are more resilient metrics than the individual IYCF indicators for predicting and monitoring progress in child nutritional health status, as postulated [
25,
34,
39,
40]. Also, in consonance with the findings of this study, none of the individual IYCF indicators was associated with LAZ in recent studies conducted in Cambodia and southern Ethiopia [
25,
41]. The fact that the general improvement in the individual IYCF indicators in the Northern Region found in this study did not necessarily show a corresponding improvement in the nutritional status of the children, is suggestive of a weak association between IYCF indicators and undernutrition (stunting and wasting). The predictive utility of the IYCF indicators individually for stunting and wasting is thus limited. This buttresses the observation that, although a prolonged state of limited or inadequate food intake has been implicated as a major cause of stunting, some emerging evidence now suggests that stunting could continue to prevail even in food-secure households (nutritionally negative deviant children) or vice versa in food-insecure households (nutritionally positive deviant children). This suggests that, among other things, sub-optimal food utilisation, in synergy with other proximal and intermediate factors, may potentially play a significant role in children’s nutritional status [
42,
43,
44].
Intake of foods rich in iron has been reported as a significant determinant of anaemia, but not often for stunting among children in DCs [
45,
46,
47]. Kubuga et al., in a 12-week, community-based feeding trial study in northern Ghana, found the consumption of iron-rich native
Hibiscus sabdariffa leaf meal to be significantly associated with anaemia in women of reproductive age (15–49 years) and had a protective effect against stunting among toddlers [
48]. This observation may be of interest and relevance in developing dual-purpose nutritional intervention programmes aimed at addressing both stunting and anaemia.
The finding about age as a significant risk factor for stunting is consistent with similar nutritional studies in northern Ghana and other DCs [
13,
15,
49]. Compared to children in the youngest age range (6–11 months), older children especially had a significantly higher risk of stunting. During this period, complementary feeding tends to be compromised, leading to adverse child anthropometric growth, while infants aged 6–11 months appear to benefit from the nutrient-rich and protective effect of breastfeeding. Maternal height has been reported as significantly associated with stunting by studies in similar settings [
15,
38]. The epigenetic influence of maternal height tends to underscore the role of genes and the environment on phenotypic expressions in children, even though the WHO child growth standards depict that all children follow similar optimal growth patterns, irrespective of their race, type of feeding, wealth or social status, among other factors, under similar environmental circumstances all around the world [
16,
50].
The covariates or risk factors that were significantly associated with wasting after adjusting for confounders were child gender, child age group and source of power for lighting the household (
Table 7). Acute undernutrition (wasting) tends to be relatively more pronounced in male children, probably due to diarrhoeal diseases arising from the greater tendencies of male children to explore their environment as they begin to crawl, toddle and/or walk [
14,
15,
51,
52]. Females are also genetically more resilient to stressors than males in early life [
53]. Child age group was associated with wasting, as reported in a similar study in northern Ghana [
54]. Young children, upon commencement of complementary feeding, are more acutely vulnerable to diarrhoeal infections and acute food inadequacies than other children much younger or older. Besides being introduced to complementary meals, breastfeeding provides babies (much younger children) with the required nourishment and protection against infections, whilst the relatively older children can better ingest the often more nutritious household meals than children in the intermediate age-groups (6–11 and 12–17 months). More often than not, these family meals will be more nutritious than the often monotonous cereal based-only porridge fed to young children [
54]. Households without access to electricity are likely to have limited capacity to safely and properly preserve food ingredients and unused food meant for their children, thus increasing children’s vulnerability to unsafe food intake.
4.5. Recommended Public Health Interventions to Address Child Stunting and Wasting
Stunting is reflective of chronic undernutrition, and thus the risk factors identified in this study inform the suggested interventions accordingly. The evidence that consumption of iron-rich foods makes a difference to stunting gives an indication of the need for some micronutrients. Nutritional interventions that improve the intake of iron-rich foods could serve the dual purpose of reducing both child stunting and anaemia. Increasing dietary diversity is a potentially viable approach to reduce the burden of stunting among young children, given that dietary diversity scores are widely acceptable predictors of the adequacy of micronutrient intake [
55,
56,
57].
While further prospective studies are needed to determine the effects of feeding practices on linear growth, food-based interventions (including diversifying food production, dietary supplementation, dietary diversification and biofortification) have great potential to ensure food and nutrition security, combat micronutrient deficiencies, improve dietary quality and raise levels of nutrition, especially among vulnerable groups, which include children under five years of age [
58]. In addition, awareness-building and communication programmes promoting livelihood and social behaviour change that support the increased production and consumption of nutrient-rich foods (such as green leafy vegetables) are needed to bring about positive behavioural changes relating to food and nutrition.
Furthermore, interventions that address dietary diversity, such as unconditional cash transfers and the community-based promotion of improved infant and young child feeding practices, should be pursued, together with efforts to reduce the burden of infectious diseases, such as seasonal malaria chemoprevention. Practices that promote the consumption of blended staples with other locally available protein-rich food sources should be encouraged to improve the quality of child food intake.
Strong advocacy by local, national, and international stakeholders who are engaged in sustainable human capital development in developing and emerging nations is recommended in order to improve children’s diets during the first two years of life, as a priority area. The government of Ghana needs to ensure that its food and nutrition security policies are locally relevant and factually appropriate but aligned with internationally agreed-upon recommendations in order to protect, promote and support age-appropriate complementary feeding practices for infants and young children. The government and its development partners, including nutrition and health-related non-governmental organisations (NGOs), need to better coordinate efforts in order to implement multisectoral, large-scale, evidence-based programmes for the promotion and support of improved complementary feeding for children aged 6–23 months in the Northern Region of Ghana.
Because wasting is reflective of acute undernutrition, it requires situational interventions. Nutrition-sensitive interventions, such as child immunization and vaccination programmes and promotion of appropriate water, sanitation and hygiene (WASH) practices to reduce the burden of infectious diseases, policies and programmes targeted at improving agriculture and food security (the availability and/or supply of food, purchasing power and optimal utilization of limited food resources) in households, should be implemented and scaled up. These could more sustainably address the limitations inherent in the overreliance on food donation and micronutrient supplementation programmes in northern Ghana to address the undesirable trend of acute nutritional status.
Nutrition-sensitive interventions, such as behaviour change education (BCE) targeted at improving nutrition-specific IYCF practices, should be continually carried out. Feeding and caregiving aimed especially at male children to minimise child morbidity, and nutritional education targeted at dietary diversification to improve the intake of nutrient-rich meals, should be adopted and scaled up during periods of food shortages or famine. The timing of such interventions should coincide with the period of highest food insecurity and highest vulnerability to infectious diseases, which is usually at the onset of the major planting season (May–August). Additionally, nutrition-specific interventions such as the production and distribution of special nutrient-rich foods, ready-to use therapeutic foods (RUTFs) or formulated meals from local food ingredients and the prompt treatment of sick children in resource-poor settings such as northern Ghana could help address acute undernutrition.
The influence of child gender and maternal height on undernutrition remains poorly understood and warrants further research. Alongside the conventionally known and widely researched determinants of undernutrition in developing countries (DCs), other potentially novel risk factors of undernutrition should also be investigated to ascertain their possible associations with the nutritional health of children in northern Ghana and other similar settings in DCs [
59,
60]. An example is the effects of traditional cereal processing methods (TCPMs) such as soaking, germination, wet-milling, dry-milling, dehulling, storage and fermentation on the nutritive and non-nutritive constituents of cereal ingredients (flour and dough) used for complementary meal preparation. More research into the food utilization dimension of household food security (
Figure S2) including some potentially unhelpful or unhealthful habitual food resource utilization practices (HFRUPs) such as some religious, traditional or indigenous food preparation methods may be very helpful [
60]. The influence of cultural beliefs and practices relating to child feeding and caregiving as potential risk factors for undernutrition could also be explored. Mycotoxin exposure as a potential risk factor for child growth impairment has been under investigation in some DCs for the last decade and should also be researched further in the northern regions of Ghana [
43,
61,
62,
63,
64].