After birth and after obtaining informed consent from parents, data of newborns of birth weight < 2500 g were collected on a standardized data collection sheet. They were then followed by pediatricians at a regular interval (1 month, 3 months, 6 months, 9 months and 12 months). We collected mothers’ socio-demographic data, antenatal data, the data of the pregnancy and delivery, newborns’ clinical parameters and during the first year of life. Psychomotor development was assessed at 12 months of life using the revised Brunet-Lézine scale of 150 items [12], which is a development test for newborns up to 30 months. It is easily done during a consultation and makes it possible to assess development in 4 areas, namely motor skills or posture, occulo-motor coordination, language and socialization. In our study, this was performed by a psychomotrician in the infant's environment (at home) and assessed the 4 domains. Thus, the various developmental quotients were normal if they were above 70. Below 70, psychomotricity was necessary.

The data was recorded on Epi Info 7.2.2. We performed a descriptive analysis to determine sample characteristics and estimated the prevalence of low birth weight newborns. Quantitative data were expressed as average and standard deviation for gestational age, age of mothers, and anthropometric parameters of newborns. We calculated the median for the length of hospitalization and documented the age at death. We compared medians by the Mann-Whitney test, averages by the Kruskal-Wallis test or the Welch test when comparing by sex, because the conditions for homoscedasticity were not met between the groups to be compared. Qualitative data were expressed in frequency, and to compare them we used the chi-square test of independence. We created univariate logistic regression analyses to measure the relationship between children’s and parents’ parameters with low birth weight. The variables were chosen for their statistical significance in univariate models and relevance based on the literature. We assessed the relationships between psychomotor development scores, birth weight, and gestational age by calculating the Pearson correlation coefficient.

The threshold of statistical significance was set at α < 0.05. The analysis was performed using CDC Epi Info 7.2.2, MS Excel, and the online statistical analysis software P-value (https://www.pvalue.io/fr/) which also allowed us to obtain the graphs of the correlation lines.

During the study period, 1260 low birth weight newborns were recorded at UHCME-JEF. The use of non-inclusion criteria allowed us to include 300 patients in our study. Among them, 81 (27.0%) were stillborn and 219 (73.0%) were monitored up until the age of 12 months.

These newborns were delivered to 197 mothers. The age of mothers ranged from 15 years to 44 years with an average of 27.7 ± 6.2 years. The median age was 28, with Q1 at 23 and Q3 at 31. The proportion of single mothers was 48.2% (n = 95) and with no occupation in 32.5% (n = 64) of cases. In 65.5% (n = 129) of cases, the number of antenatal contacts was less than or equal to 4 and in 85.4% (n = 176) of cases, antenatal care was carried out by midwives. 35.5% (n = 70) of studied mothers were primiparous and 54.8% (n = 108) poor. The main intercurrent pathologies of pregnancy were malaria (19.3%, n = 38), hypertension (16.8%, n = 33) and vaginal infection (12.7%, n = 25). Pregnancy was unique in 89.3% of cases (n = 176). In 79.2% (n = 156) of cases, delivery was vaginal.

Table 1 shows the characteristics of low birth weight neonates. The mean gestational age was 34 ± 12 weeks, with extremes ranging from 23 to 41 weeks. The median gestational age was 34 weeks with a Q1 of 32 weeks and a Q3 of 36 weeks. In 75.8% of cases, newborns were premature. The mean birth weight was 2008.6 g ± 215.3 g with a range of 630 g to 2480 g. LBW newborns were observed in 84.0% (n = 184) of cases. The mean cranial perimeter was 30.7 cm ± 3.5 cm, with extremes ranging from 24 to 35 cm. The mean length was 43.1 cm ± 4.2 cm, with extremes ranging from 27 to 58 cm. The sex ratio was 0.8. The intra-uterine growth retardation (birth weight< 10^th percentile according to the gestational age) was observed in 13.2% of the cases (n = 29). Of these, 23 showed disharmonious intra-uterine growth retardation and 6 showed harmonious intra-uterine growth retardation. Macrosomia (weight > 90^th percentile by gestational age) was observed in 2.3% of cases (n = 5).

· Distribution of patients according to hospitalization at birth

At birth, 44.3% (n = 97/219) of patients were hospitalized in the neonatal medicine department. The 3 main reasons for hospitalization were neonatal infection (27.8%, n = 27), prematurity (22.7%, n = 22) and perinatal asphyxia (20.6%, n = 20). They were followed by respiratory distress (18.6%, n = 18), intrauterine growth restriction (5.2%, n = 5), neonatal jaundice (4.1%, n = 4) and ichthyosis (1%, n = 1). The average length of a hospital stay was 9.6 ± 8.2 days, with extremes ranging from 2 to 60 days. The median was 7 days with Q1 at 4 days and Q3 at 15 days.

· Childhood pathologies presented during the first year of life

In 61.6% (n = 135) of cases, a childhood illness was observed. Respiratory

pathologies (rhinopharyngitis, bronchopneumopathy and pneumonia) had a higher proportion than other pathologies in all age groups with respectively 37.4% (n = 82) at 1 month, 13.7% (n = 30) at 3 months, 12.8% (n = 28) at 6 months, 11.4% (n = 25) at 9 months and 8.7% (n = 19) at 12 months. At 3 months of age, the proportion of digestive pathologies (constipation, gastroenteritis, dietary error, regurgitation and wrong route) predominated (20.5%, n = 45). The overall proportion of pathologies encountered during growth decreased with age. At 12 months, 7762% of infants presented no pathology at the medical control.

All patients had a psychomotor development quotient > 110 at 12 months of life. Postural development quotients, language development quotient and coordination development quotient were normal (between 110 - 70) in 75.4%, 99.1% and 68% of the cases respectively at 12 months of age.

· Language development quotient of patients at 12 months of life

There was no association between language development quotient (LDQ) and maternal parameters or with infant pathologies contacted during the 12 months of life. Language development quotient (LDQ) in neonates with perinatal asphyxia at birth was on average 82.5 (±11.2). For those who had not been diagnosed with perinatal asphyxia, it was on average 87.1 (±7.9). There was an association between perinatal asphyxia and language development quotient with a p = 0.019 (Figure 1). The correlation coefficient between language development quotient and birth weight was r = 0.15 with p = 0.024. There is a correlation between birth weight and language development. Patients with very low birth weight had poorer language development than those with low birth weight (Figure 2).

· Coordination development quotient (CDQ) of patients at 12 months of life

There was no association between language development quotient (LDQ) and maternal parameters, nor with infant pathologies contacted during the 12 months of life. In patients, the correlation coefficient between Coordination development quotient (CDQ) and birth weight was r = 0.15 with a p = 0.026. There was an association between CDQ and birth weight. Very low birth weight patients had worse coordination development than low birth weight patients (Figure 3). There was no association between perinatal asphyxia and CDQ.

· Postural development quotient (PDQ) of patients at 12 months of life

There was no association between developmental quotient and maternal parameters, nor with infant pathologies contacted during the 12 months of life. In patients with perinatal asphyxia at birth, the postural development quotient

(PDQ) averaged 71.5 (±17.8). In those without perinatal asphyxia, the average was 79.7 ± 11.8 with a p = 0.026. There was an association between Postural development quotient (PDQ) and perinatal asphyxia (Figure 4). The correlation coefficient between PDQ and birth weight was r = 0.15 with a p = 0.015. There was an association between postural development quotient and birth weight. Very low birth weight patients had worse postural development than low birth weight patients (Figure 5).

The context of COVID-19 pandemic as well as the health measures taken has created certain obstacles to the optimal conduct of this survey, in particular the

unique nature, the lack of patient attendance during outpatient follow-up, and the low adherence of parents to the study. Nevertheless, the results obtained have made it possible to identify useful data, which can contribute to the improvement of the management of newborns of low birth weight in our health centers.

The frequency of low birth weight at UHCME-JEF during the study period was 14.0%. This frequency is superposable on that observed in Gabon globally in 2004 (14.0%) [13], close to that estimated in sub-Saharan Africa (13.0%) [10] and those reported by Faye in Dakar, Bwana Kangulu in Kamina, Tshinzobe in Kinshasa and Muchemi in Olkalouqui are respectively 14.8%, 14.3%, 13.3% and 12.3% [14] [15] [16] [17]. It is lower than those found by Alemu in Ethiopia (18.0%) [18] and Chiabi in Cameroon (20.7%) [19]. These different prevalences of low birth weight all remain high. This finding may be because the UHCME-JEF is a reference facility for the management of pathological pregnancies and sick newborns and on the other hand by the high rate of intercurrent pathologies in pregnancy (56.4%), dominated by infectious and vascular pathologies that are known as providers of low birth weight [10]. These intercurrent pathologies in pregnancy have also been reported in the studies of Faye and Chiabi [14] [19]. Also, it is well established that it is in developing countries that the rate of low birth weight is most found [20]. This is probably due to the high prevalence of prematurity and intercurrent pathology to pregnancy that they record.

The sex ratio was 0.8. This result is identical to those of Faye. et al. in Senegal, Bwana Kangulu et al. in DRC, Chiabi et al. in Cameroon, and Hassoune et al. in Morocco, which find a female predominance [14] [15] [19] [21]. Bandyopadhyay et al. in India find a male predominance (54.2%) with no impact on the fate of newborns of LBW [22]. The average age of newborns and the proportion of preterm newborns in our study are superimposed on those found by Faye, Chiabi and Bandyopadhyay who found a rate of preterm newborns of 85.6%, 79.6% and 78.2% respectively with an average gestational age of 34 weeks and a higher proportion of average prematurity [14] [19] [22]. This is not surprising because prematurity is one of the main causes of low birth weight [7] [10]. This high rate of average prematurity also explains the fact that the average birth weight in our study was 2008.6 ± 215.2 g with a higher proportion of newborns with LBW. This observation is also made by Faye, Chiabi and Bandyopadhyay [14] [19] [22].

At 12 months of age, in general, the majority of low birth weight infants had normal developmental quotients, between 70 - 110 for PDQ (75.3%), (68.0%) for DCQ and (99.1%) the development quotient of language. These results are similar to those of Sylla M et al. who found 71.2% and 90% respectively for language and motor skills [23]. We also observed that there is an association between psychomotor development and birth weight. Indeed, there is a significant difference in the different partial development quotients (posture, language and coordination) between very low birth weight and low birth weight newborns with a p = 0.015, p = 0.024 and p = 0.026 respectively. This is similar to the results of Savadogo LGB et al. who observed that the psychomotor development of low birth weight infants was significantly slower than that of very low birth weight infants [24]. This may be explained by the fact that the majority of low birth weight newborns are preterm and nutritional status has been shown to have a significant impact on the psychomotor development of the infant [25].

Perinatal asphyxia was the main factor of poor development in posture and language in our study. It is well established that one of the most severe complications of perinatal asphyxia is brain damage and therefore explains this observation. This observation was also made by Stacey et Al in Canada where asphyxia was the cause of global developmental retardation in 55% of cases [26]. Similarly, Kedy Koum in Cameroon observed in a population of children with neonatal asphyxia behavioral abnormalities (8.2%), prehension abnormalities (8.2%), gestural/motor abnormalities (14.3%) and language abnormalities (26.5%), ass well as tone disorders (10.2%) and delays in language acquisition (27.0%), and walking (14.0%) [27].