This was a monocentric retrospective cohort and analytical study. The study was carried out in the pediatric department of the Essos Hospital Centre in Yaoundé within the Approved Treatment Center for antiretroviral therapy (ACT). The ACT is a specialized and a referral Centre in Prevention of mother-to-child transmission (PMTCT) of HIV and for care of children living with HIV. This study was conducted for 3 months, from March to May 30, 2018 and covered a period of 12 years from 2005 to 2017. Patients enrolled were responding to the following criteria: being HIV-infected children aged 15 years or less, on ART and follow up at Essos Hospital Centre, Files with key missing data were excluded. The children included were then divided into 2 groups. Group 1 started ART before 2 years and Group 2 started treatment after the second birthday.

After obtaining ethical clearance, under the number 2018/04/CE-CHE we began the recruitment. In absence of exclusion criteria, written informed consent was obtained. The data was collected during a face-to-face interview by the research investigator. The main variables were 1) sociodemographic; 2) therapeutic and clinical according to the 2007 WHO clinical staging. With regards to anthropometric variables relating to growth, we evaluated weight and height. These parameters were reported as percentiles (weight-for-age; height-for-age) using the Center for Disease Control and Prevention growth charts. The results obtained in percentiles were converted into standard deviation; the 3 and 97 percentiles corresponding respectively to −2SD and +2SD [13] ; 3) Biological: CD4: we used the WHO classification of immunological status. Viral load: the virological status was assessed as undetectable at a threshold < 60 copies/ml.

Child: any individual under the age of 15; Stunted: height-for-age index less than −2SD. Underweight: weight-for-age index below −2SD. “Normal” height growth: height-for-age index greater than −2SD. “Normal” weight growth: weight-for-age index greater than −2SD; Group 1: children initiated on ART before the age of 2 years ; Group 2: children initiated on ART after the 2^nd birthday.

Data analysis was done using Excel, SPSS and Stata 12 software. The qualitative variables were expressed in the form of numbers and proportions; as for the quantitative variables, the distributions were represented by their mean. The comparison between the proportions was made using the chi-square test. To determine the association between the W/A, H/A greater than −2SD and the independent variables, we used logistic regression. All the variables associated with a normal weight and height in univariate analysis were introduced into the same logistic regression model to find the determinants of a W/A and H/A > −2SD. The significance level was set at 5%.

This study protocol was submitted to the institutional review board of EHC and the ethical was approved under the number 2018-04/CE-CHE. In addition, we obtained a research authorization from the administrative authorities of the Yaoundé EHC. All eligible persons have been explicitly built on the interest of our study. Participant approval was obtained and recorded in an informed consent form.

In our study, we targeted 115 subjects from the available files. Nine of them were unusable. Of the remaining 106, we definitively recruited 90 subjects because of refusal to participate (n = 6), ineffective phone numbers (N = 5) and deaths (4).

We included a total of 90 subjects including 46 (51.1%) girls and 44 (48.9%) boys with a male/female sex ratio of 0.95. 52% of the children were initiated before the age of 2; the average age at initiation was 10 months. For children initiated after the age of 2 years (48%), the age at initiation varied between 2 and 13 years, with an average age of 6 years. At the time of the study the age of the children varied between 2 and 15 years with a median of 10 years, with 25% lived with both parents. Most of the children were in school. Regarding the status of the parents, of the 56 mothers whose status could be collected, 2 were HIV-negative (3.6%), in the case of the fathers, 40% were HIV-negative.

For children initiated before the age of 2 years Group 1, we find that: 52% had a weight deficit at ART start against 4% at the time of the study. Thus, these children almost all regained their weight on ART. On the other hand, 15/44 (34%) were stunted at ART start versus 4/44 (9%). Thus, after initiation of ART, almost all children have caught-up in height.

Of the 90 children at the time of our study, 81% had good compliance and 19.3% had interrupted their treatment, 33.7% knew their “HIV” status at the state of complete disclosure 24.4% had never been hospitalized for more than 24 hours and 54.8% were on cotrimoxazole at the time of the study ( Table 2 ).

In view of the results of the univariate analysis, starting ART before the age of 2 years did not influence the weight of the children in our study ( Table 2 ). In addition, children who did not interrupt their treatment had a greater chance (OR (95% CI); 2.856 (1.84 - 4.42)) of having a normal weight for age compared to those who had interrupted, regardless of the age of initiation ( Table 3 ).

The interruption of treatment and the start of ART before the age of 2 years were the main factors influencing height of the children of our entire study population ( Table 4 and Table 5 ). The multivariate analysis reveals that: 1) children who do not interrupt their treatment were more likely (OR (95% CI; 2.72 (1.78 - 4.15))) to have a normal height than those who interrupt their treatment regardless of is the age of initiation. 2) those who started treatment after the age of 2 years had lower chance (OR (95% CI; 0.078 (0.001 - 0.45))) of reaching a normal height than those who started treatment before the age of 2 years.

However, the effect of the interruption on growth differs according to the authors; some highlight its interest in optimizing treatment: Coton et al. [21] in a study found that interrupting ART over time after initiation at an early age showed good clinical and immunological responses after resumption of treatment compared to the deferred treatment group. Penezzato et al. in 2014 during a study on the optimization of ART in HIV-infected children found that children who interrupted their treatment had similar growth than those who had continuous treatment [22] . These differences could be explained by the duration of the interruption which length varies in various studies from less than 3months to 33 weeks [21] . Indeed, Dalton et al. found that a “short” interruption duration did not compromise growth [23] .

The age of initiation to treatment did not influence the weight of the children in our study. Our results differ from those found by Diniz et al. in Brazil [24] who found that young age at initiation was a predictor of greater weight catch-up. Similarly, Nachman et al. [8] in a study similar to ours reported that children under 2 years of age were more likely to catch up on their weight compared to older ones. Some authors in the European context claim that it would take up to 5 years of exposure to ART to catch up to normal height [25] .