This was a longitudinal cohort study of children (12 - 59 months) admitted to INSE HN Donka for moderate malnutrition from October 2023 to June 2024. This study included a phase of food sampling and quality control, targeted recruitment of participants, a nutritional intervention, and regular monitoring of anthropometric parameters. It aimed to evaluate the impact of a food combination composed of Parkia biglobosa, honey, and powdered milk.

The first phase of the study was carried out at the laboratory of the National Office of Quality Control, located in Matoto (Conakry). This office is a public service with a scientific and technical vocation, placed under the supervision of the Ministry of Commerce and Industry. Its mission is to control the application of regulations relating to the quality of food consumption needs in the Republic of Guinea.

The second phase of the study took place in the neonatology department of the Institute of Nutrition and Child Health (INSE), a public institution founded in 1989 and dedicated to care, research, and training. The INSE comprises several departments: an administrative department, a neonatology department (with maternal counseling, vaccination, and physiotherapy units), a nutrition department (with renutrition, nutritional education, and milk preparation units), a research and training department, as well as a laboratory and pharmacy department (grouping together various specialized units).

The study population consisted of children aged 12 to 59 months who came for consultation at INSE HN Donka, were diagnosed with moderate malnutrition, and whose parents/guardians gave informed consent.

For the selection of children, the following criteria were observed:

The inclusion criteria were as follows: children aged 12 to 59 months with moderate acute malnutrition (Z-score W/H or W/A between −2 and −3 standard deviations, or mid-upper arm circumference between 11.5 cm and 12.5 cm) [14] , and voluntary parents/guardians (informed consent).

Non-inclusion criteria included children under 12 months or over 59 months, those with severe malnutrition or no malnutrition, and children with acute conditions requiring immediate hospitalization.

Children meeting the inclusion criteria and admitted to INSE HN Donka between October 2023 and June 2024 were consecutively included using a purposive, exhaustive, hospital-based convenience sampling approach. This approach is regularly used in clinical settings to cover all available cases without resorting to formal random selection. A total of 31 children were enrolled during the study period.

Before any use, each food, namely P. biglobosa flour, honey, and Nido powdered milk, was subjected to laboratory analyses by the National Office of Quality Control (ONCQ) located in Conakry to guarantee their safety and quality. The P. biglobosa flour and honey were purchased in the province of Kankan and stored hermetically at room temperature. Boxes of Nestlé powdered milk (800 g) were obtained from wholesalers in Conakry.

Sensory analyses were carried out according to the standard [15] to evaluate color, odor, flavor, and texture. Simultaneously, physicochemical analyses were conducted, including pH, humidity [16] , total acidity [17] , Brix degree [18] , and ash rate [19] of the ingredients. Finally, microbiological analyses were performed, including enumeration of Total Aerobic Mesophilic Flora [20] , Escherichia coli [21] , fecal streptococci [22] , as well as yeasts and molds [23] .

This was a sequential process that began by mixing 3 tablespoons of P. biglobosa flour in a container with 50 ml of mineral water, then boiling for approximately 2 minutes. After cooling, 1 tablespoon of Nido milk was added, stirring evenly before also adding 1 tablespoon of honey. Finally, the remaining 40 ml of water was added while continuing to mix with a spoon to obtain a smooth paste.

Data collection was structured around several key stages.

The recorded cases of moderate acute malnutrition were subjected to the P. biglobosa flour-honey-Nido powdered milk combination during their care period. The entire care period was thirteen (13) weeks. Each week, the child received the combination, followed by anthropometric measurements. Semi-structured interviews with parents/guardians made it possible to assess the acceptability of the formula and any challenges related to its administration. The records documented the quantity of formula consumed and the frequency of intake.

Weight gain or net weight gain

This is the weight increase from the first week to the thirteenth week (quantitative variable).

This is the number of times you take a food supplement per day.

Anthropometric data were collected on the first day of the study and at regular intervals (weekly) to assess the children’s growth. These measurements, carried out by trained personnel, were standardized to ensure their reliability [24] . These included mid-upper arm circumference (MUAC) (measured on the left arm with a non-stretchable tape to the nearest 0.1 cm), weight (measured with an electronic pediatric scale calibrated to the nearest 100 g), and height (measured with a height rod to the nearest 0.1 cm). For analysis, these data were then converted into Z-scores (weight-for-height, height-for-age, weight-for-age, and mid-upper arm circumference-for-age) based on WHO growth references [25] .

Data were entered and analyzed using SPSS version 22 software. Descriptive analyses focused on quantitative variables (weight, height) expressed as mean ± standard deviation and on qualitative variables as numbers and percentages. The weight gain variable was obtained by calculating the difference (Weight_S13 - Weight_S1).

The normality of the distributions was verified by the Shapiro-Wilk test, and the homogeneity of variances (sphericity) by Mauchly’s test.

To compare the evolution of anthropometric parameters between the first and last week (W1 to W13) of follow-up, a repeated measures ANOVA (factor “time”) was applied when the parametric conditions were satisfied (normality and sphericity, with Greenhouse-Geisser correction if necessary). In case of non-normality or violation of the parametric assumptions, a non-parametric alternative was used: the Friedman test, the non-parametric equivalent to repeated measures ANOVA [26] . The statistical significance threshold was set at 0.05 (p < 0.05) [27] . All analyses were two-tailed.

Ethical approval was obtained from the Institute Ethics Committee of the National Public Health Agency. Informed consent from parents/guardians was a prerequisite for each child’s participation. Data confidentiality was strictly respected.

The most represented age group was 30 - 39 months (32.3%), followed by 20 - 29 months (29.0%). Children aged 50 - 59 months were the least numerous (3.2%). Regarding gender, boys (54.8%) were slightly more numerous than girls (45.2%). Table 1 shows the distribution of children (12 - 59 months) according to age group and gender.

Table 1. Distribution of children (12 - 59 months) by age group (months) and sex from October 2023 to June 2024 INSE HN Donka.

In terms of origin, the majority of mothers resided in Matoto (25.8%), followed by Kaloum (22.6%) and Ratoma (19.4%). Maneah and Dixinn were the least reported origins (3.2% and 6.5%, respectively).

Marital status reveals that the vast majority of mothers were married (87.1%), which suggests a stable family structure for most of the children in the study. Single (9.7%) and divorced (3.2%) mothers were in the minority.

As for the mothers’ occupation, shopkeepers represented the largest proportion (51.6%), which is an important socioeconomic factor to consider. Accountants and seamstresses had similar proportions (12.9% each), while hairdressers were the least represented (6.5%). “Other” occupations (16.1%) included civil servants and vendors. Table 2 describes the sociodemographic characteristics of the mothers of the children studied.

Others: Civil servants (4), merchants (1).

Only 3.2% of children were between 100 - 109 cm tall. The majority of children’s heights were between 80 - 89 cm (51.6%), followed by children under 80 cm (29%).

The weight-for-height (W/H) Z-score indicated that 81.3% of children had a Z-score of −2 and 18.8% had a Z-score of −1.

Regarding mid-upper arm circumference (MUAC), 87.5% of children had a MUAC between 11.5 and 12.4 cm, and 12.5% had a MUAC greater than 12.6 cm.

All children therefore met at least one of the WHO criteria for moderate acute malnutrition (MUAC 11.5 - 12.5 cm or −3 ≤ WHZ < −2) at baseline. Table 3 details the anthropometric characteristics of the children.

The semi-structured interviews conducted with mothers revealed high acceptability of the Néré-honey-milk mixture. Caregivers reported good palatability, improved appetite and energy among children, and willingness to continue preparing the mixture at home after the study. No adverse effects were reported.

Analysis of children’s weight change over the 13-week follow-up was performed using the non-parametric Friedman test, appropriate for paired data with repeated measures. This test allowed comparison of weights recorded at different follow-up time points (weeks 1, 7, and 13). The results revealed a statistically significant difference in weights measured over time (Friedman’s χ² = 60.06, p < 0.001).

The two-way analysis of variance by Friedman rank-ordering of related samples is illustrated in Figure 1 , which presents the distribution of children’s weights over the 13 weeks of follow-up.

The analysis of weight change, carried out using the non-parametric Friedman test, showed a statistically significant difference (p < 0.001) in the weights measured at weeks 1, 7, and 13. This result, with a chi-square value of 60.06, indicates a significant improvement in the children’s weight over time, which is a positive result in the follow-up.

Figure 2 explores the relationship between weight gain and children’s origin.

There is no statistically significant difference according to the Kruskal-Wallis test between weight gain and the origin of the children (p = 0.600) α = 0.05 CI 95.

Again, the analysis (p = 0.131) reveals no statistically significant difference between children’s weight gain and their mothers’ marital status. Figure 3 examines the relationship between children’s weight gain and mothers’ marital status.

For age, the Kruskal-Wallis test (Chi-square = 7.686, p = 0.104) indicates no significant association between age and weight gain, although a trend can be observed with a higher mean rank for younger children (10 - 19 months).

Regarding height, there was a statistically significant association (Chi-square = 10.871, p = 0.012) with weight gain. The shortest children (<80 cm) had a higher mean weight gain rank (24.33) compared to other height groups, which could suggest that the shortest children at the start of follow-up had more pronounced weight gain.

Finally, MUAC (Chi-square = 1.47, p = 0.225) was not significantly associated with weight gain, suggesting that initial nutritional status (as assessed by MUAC) did not directly influence the weight gain observed during follow-up.

Table 4 analyzes the relationship between weight gain and children’s age and height, as well as their mid-upper arm circumference.

Adherence score: The mean adherence score was 8.4 ± 1.2 out of 10, reflecting a high level of compliance with the feeding protocol. No significant correlation was found between adherence score and weight gain (p = 0.24), although a positive trend was observed among children with the highest adherence.