This was a prospective, longitudinal, descriptive cohort study. It ran from 1 May 2022 to 03 June 2024, i.e., a period of two (2) years. The study took place in the municipality of Libreville with an estimated population of 703,939 and an area of 654.42 km² [16] . The study setting was the neonatal medicine department of the CHUME-FJE in Libreville. This is a level 3 hospital dedicated to maternal and child health. The study population consisted of all macrosomic newborns born at the CHUME-FJE during the study period. We included newborns with a birth weight (BW) greater than the 90th percentile of the reference curves whatever the term; patients for whom we obtained informed consent from the parents. We did not include stillbirths, neonates whose parents did not give consent, neonates with excess extracellular fluid (such as hydrocephalus and/or hydrops), neonates with sacro-coccygeal malformations or tumours, or congenital neck cysts. Excluded neonates were those lost to follow-up and those whose parents withdrew after inclusion.

Data were collected from delivery room registers, maternity ward medical records, neonatal medicine department medical records and outpatient clinic records. They were then collected on a pre-established standardised form completed during outpatient follow-up at regular intervals (1 month, 3 months, 6 months, 9 months, 12 months). Our sample size for statistically significant analysis (N = 57) was determined using the Schwartz formula with a prevalence of 3.85% [10] . The number of macrosomic newborns was 118. We included 100 newborns meeting the criteria. Of these, 60 patients were included in the study, 28 of whom refused to participate and 12 of whom were lost to follow-up. The excluded newborns (18) had malformations affecting weight.

In the newborn: gestational age, sex, route of delivery, Apgar score, anthropometric parameters, feeding of the newborn, kinetics of changes in anthropometric parameters.

To avoid bias and ensure the validity of our results, we used the following methodology. To avoid selection bias, each patient was recruited when birth weight was above the 90th percentile of the Lubschenco curves. To avoid information bias, patients were followed-up in the presence of the same paediatrician. Weight was measured using the same device (SOEHNLE brand). The same measuring tape was used to measure height, and the same tape measure was used to measure head circumference. Finally, to avoid confounding bias, we excluded from the study newborns with malformations (hydrocephalus, sacrococcygeal tumors, congenital neck cysts, hydrops, etc.). In fact, these anomalies will lead to an increase in weight, without true growth.

The study began in the delivery room. After obtaining informed consent, the newborn was recruited as soon as the weight relative to the Lubchenco growth curve was above the 90th percentile. The variables were collected on a pre-established standardised form. After discharge from hospital, outpatient follow-up was carried out at regular intervals (1 month, 3 months, 6 months, 9 months, 12 months) in the presence of the outpatient paediatrician. It was based on a physical examination of the newborns, monitoring of the growth curve (control of anthropometric parameters: weight, head circumference, height), and the search for any pathologies occurring during growth.

During the follow-up, anthropometric measurements: height, weight and head circumference were taken in order to assess the weight and height development of the newborns in relation to their age. At each visit with the paediatrician, we measured weight using a baby scale (SOEHNLE), cranial perimeter using a tape measure, and height using a measuring tape. These various anthropometric parameters were then collected and projected onto growth curves using WHO Anthro software in order to compare each child with the norms for his or her age and sex [17] - [19] . The Z-score thresholds were chosen because they are the growth charts used in the department from 1 month of age for staturo-ponderal growth and the nutritional status of children.

Statistical analysis was performed using R software version 3.4. Quantitative variables were expressed as mean and standard deviation, and qualitative variables as number and frequency. The various anthropometric parameters collected during follow-up were projected onto growth curves using WHO Anthro software in order to compare each child with the WHO norms for age and sex [17] - [19] .

We obtained all the necessary administrative authorisations to carry out our work at the centre. The informed consent of the parents had been obtained. Confidentiality and anonymity were respected throughout our work.

At birth, 18.3% of newborns were obese. At 1 and 3 months, growth was normal in 80.0% and 78.3% respectively. At 3 and 6 months, overweight infants accounted for 20% and 31.7% respectively. At 9 months, the weight growth curve was declining (69.3%). At 9 and 12 months, infants remained obese at 28.3% and 23.3% respectively ( Table 2 ). At the age of 12 months, the infants had a weight growth curve above the WHO reference curve ( Figure 1 ).

From birth to 12 months, infants showed normal statural growth. At 3, 6 and 9 months of age, stunted statural growth was observed in 5.0% (n = 3), 13.4% (n = 8) and 11.7% (n = 8) respectively ( Table 2 ). At 12 months, statural growth was normal ( Figure 2 ).

At birth and at 1 month, the newborns were obese (15.0% and 23.3%). At 3, 6 and 9 months of age, staturo-weight growth was normal (76.7%, 63.3% and 65.0% respectively). Obesity was found in 21.7%, 36.7% and 30.0% of cases respectively. At 9 and 12 months, an increase in staturo-weight growth was observed, with a mean z-score of 1.5 ± 1.6 (median 1.5) and 2.2 ± 1.2 (median 2.1) respectively ( Table 2 ). At 12 months, 35.0% of infants remained obese ( Figure 3 ).

From birth to 12 months, infants had normal head circumference growth (78.3% and 50%). Macrocrania was observed in 20.0% and 41.5%. At 3 months of age, 15.0% (n = 9) had delayed head circumference growth, with catch-up at 12 months ( Table 2 ) ( Figure 4 ).

A number of biases need to be taken into account when validating the results obtained, so as to be cautious in interpreting them: not all eligible patients were included, although the participation rate was 60%; the monocentric nature of the study; low parental participation in the study (28 refusals). However, this study, although hospital-based, has the advantage of being carried out in a referral hospital dedicated to the care of children of all ages.

The hospital prevalence of macrosomia was 7.2%. This incidence is corroborated by most of the data in the literature, which varies between 4 and 10.0% [1] . This result is lower than that found by Ridha et al. in Tunisia (10.9%) and Bakkali et al. in Morocco (14.3%) [2] [8] . It is comparable to that found by Usta et al in Turkey (8.6%) [6] . This rate remains higher than that found in Gabon (3.89%), the DRC (5.7%) and Guinea-Conakry (5.8%) [1] [3] [13] . This difference may be explained by the type of methodology used and the level of health structure. The high rates in the Maghreb could be explained by the socio-demographic nature of this region. Indeed, the incidence of foetal macrosomia varies from one region to another and within the same region according to ethnic origin, race and environmental factors [3] . North Africa is one of the regions with the largest increase in the number of diabetic patients, mainly type 2 [20] . According to the International Diabetes Federation (IDF), projections for 2045 put the incidence of diabetes in North Africa at 110% [21] .

The mean gestational age was 39.1 ± 1.8 SA. This result is comparable to that of Usta et al. (39.2 ± 1.5) and Ridha et al. (40.2) [6] [8] . Macrosomic newborns were born at term (88.3%) and post term (8.4%). This result differs from that reported in the literature. Most authors find a positive correlation between gestational age and fetal macrosomia [1] [3] [22] . Fetal macrosomia is thought to favour prolongation of term through fetal-pelvic disproportion. It thus disturbs the spontaneous onset of labour by modifying the mechanical components [3] .

The sex ratio is 2.3. As shown in most studies, there is a clear predominance of males [1] [3] [13] [23] . Male sex has been described as a risk factor for macrosomia [9] [24] [25] . However, no argument has yet been put forward to explain this trend. On the other hand, studies have shown that male sex is a predictor of macrosomia [1] [6] .

The mean birth weight was 4081.5 g ± 391.2 g. This mean birth weight is lower than that found by Yao et al. in France, Camara et al in Mali, Diallo et al in Guinea Conakry, and Luhete et al. in the Democratic Republic of Congo, who reported respectively 4249.8 g; 4557.5 g; 4255 g and 4207.3 ± 297.8 [1] [5] [13] [23] . This difference may be linked to the inclusion criteria. In our study, we included all newborns with a birth weight greater than the 90th percentile of the reference curves. In the study carried out by the aforementioned authors, the inclusion criterion was any newborn with a birth weight greater than or equal to 4000 g. The incidence of macrosomia is reported differently according to racial, ethnic and regional differences [13] .

In our study, weight and statural growth and head circumference were assessed progressively from birth to 12 months.

Fetal macrosomia predisposes to childhood obesity. The metabolic and cardiovascular risk of this obesity is debated [26] . Macrosomia occurring in the context of maternal diabetes exposes the baby to a significantly high risk of developing early metabolic syndrome. This metabolic syndrome predisposes to cardiovascular disease in adulthood and implies the development of precursor signs of these diseases in childhood or adolescence [14] . In the Sustainable Development Goals adopted by the United Nations General Assembly in 2015, the fight against non-communicable diseases is considered to be one of the health challenges for the 2030 Agenda for Sustainable Development [27] [28] . Among the risk factors for non-communicable diseases, overweight and obesity are of particular concern and could jeopardise many of the health gains that have contributed to longer life expectancy [28] . The Global Action Plan on No communicable Diseases 2013-2020 calls for a halt to the increase in adolescent obesity. The comprehensive implementation plan for maternal, newborn and young child nutrition sets a target of no increase in the percentage of overweight children by 2025 [28] . Several factors may be involved in childhood obesity:

This is the period from conception (270 days) to two years of life (730 days). Over thirty years ago, Barker, a British epidemiologist, showed that undernutrition during pregnancy was the cause of low birth weight and increased the risk of myocardial infarction in adulthood. Following Barker’s work, a number of researchers have confirmed that the environment to which the embryo, foetus and newborn are exposed influences the onset of certain chronic diseases in adulthood, such as obesity, diabetes and high blood pressure. These studies are now grouped under the concept of DOHaD (Developmental Origins of Health and Diseases) [29] [30] . In the case of foetal macrosomia, early exposure to high concentrations of insulin and leptin leads to excessive weight gain. Diabetes, obesity and excessive gestational weight gain can “programme” trajectories of adiposity and metabolic health throughout life [30] . It is now well accepted that the first years of life are decisive for long-term health. In particular, nutritional intake during this critical period of development (the first 1000 days) appears to have long-term effects on the health of future adults [26] .

On average, BMI increases rapidly in the first year of life. It then falls, reaching a minimum around the age of 6, before rising again steadily until the end of growth. The point at which BMI reaches its minimum value before a second increase is called the ‘adiposity rebound’. A large number of studies that have examined the predictive value of adiposity rebound have shown that an early rebound is associated with subsequent overweight [31] . On average, the rebound occurs around the age of 6, but individual trajectories show that it may occur earlier or later [32] . Most children who are overweight at the start of life will return to the average after a late rebound (after the age of 6), while others will remain fat after an early rebound [32] . The age at which adiposity rebounds is much earlier in the obese, suggesting the influence of early intervening factors [32] .

This justifies the importance of monitoring macrosomic infants. Indeed, the 35.0% of infants in our study who were obese at 1 year of age need to be monitored regularly in order to prevent long-term complications.